article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
in the last few years , there have been a large number of different approaches to the dark energy enigma . as is well - known , the anisotropies of the cosmic microwave background ( cmb ) radiation @xcite and data from type ia supernovae @xcite , are well - fit by a cosmological constant , albeit a fine - tuned one , of which we have no satisfactory theoretical understanding .
another possibility is that the cosmological constant is precisely zero ( or at most subdominant ) and that a dynamical component is driving cosmic acceleration .
this component may be a manifestation of a large - scale modification of general relativity @xcite , or a new dynamical field , such as a scalar field @xcite .
it is quite natural to consider scalar fields that are non - minimally coupled to gravity @xcite .
such fields arise both in the context of string theory , and in more general theories with extra spatial dimensions .
these fields are typically of the brans - dicke ( bd ) type , and have been considered in a large variety of other cosmological contexts , such as inflation and baryogenesis @xcite .
however , there exist tight constraints on bd fields .
some of these constraints are on solar system scales , constraining the bd parameter to satisfy @xmath0 .
other constraints arise from cosmology itself .
for example , in order not to spoil the success of the big bang nucleosynthesis ( bbn ) predictions , the scalar field component should be subdominant to radiation and the scale factor must evolve nearly as @xmath1 .
this constraint is quite well accepted , and a detailed analysis of the constraints on other kinds of evolutions , such as dark matter dominated one , can be found in @xcite . in this paper
we revisit this particular issue , considering whether a bd scalar field might play a _
significant _ role in cosmic evolution during bbn , while maintaining an acceptable evolution of the scale factor .
we shall only require the bd field to drive the kinematics of the universe , making it expand _ as if _ radiation dominates .
it should be emphasized that in our model the expansion rate ( hubble function @xmath2 ) is driven by the bd field , while the scattering rates are determined by cross sections and abundances of the standard fields of the plasma ( @xmath3 , neutrinos , photons , baryons ) .
one consequence is that the relation between temperature and time becomes an adjustable parameter in this model .
provided the bd field can reproduce a standard kinematics @xmath4 , one is free e.g. to normalize the abundances of all the fields to the value they assume at the same @xmath4 in the standard model . since they do not interact with the bd field but gravitationally , the particles in the plasma can not distinguish which component is driving the evolution of the universe , and they will interact among themselves in the same way as if the bd field is not present . as a consequence ,
the production rate for the light elements is the same and the freeze - out temperatures for the reactions are unchanged .
the big difference between this model and other bd models during bbn ( see e.g. @xcite ) , is that now the bd field , i.e newton s constant , can be largely different from today s value at the time of bbn .
this is quite different from a standard approach in which the bd field is allowed to be just slightly different from today s value .
however , just after bbn , some non - trivial dynamics is required to bring the bd fields to values consistent with a standard evolution for the universe .
this is sufficient to ensure that the temperature of the universe has the usual behavior from the end of bbn up to now .
we shall see that such a scenario is possible , while remaining consistent with solar system constraints .
however , as one might expect , this interesting possibility can only arise under finely tuned conditions .
we describe the tuning required , both on the initial conditions of the scalar and on its associated potential .
to study the cosmological aspects of this theory in the framework of a homogeneous and isotropic universe , we use the friedmann , robertson - walker ( frw ) ansatz for the metric @xmath19 where @xmath20 is the scale factor and @xmath21 is conformal time , related to the usual cosmic time through @xmath22 .
we have assumed a zero curvature contribution , consistent with current data .
in the standard scenario , the universe is radiation dominated during bbn .
this requirement leads to theoretical expectations that are in reasonable agreement with observations , at least at the 2 @xmath28 level .
in particular the deuterium number density @xcite points to a value for the baryon content of the universe which is within 1-@xmath28 equal to the independent determination of this parameter extracted from the cmb power spectrum @xcite @xmath29 this result is compatible with a radiation content given by photons and three weakly interacting neutrinos / antineutrinos ( for a review of the present status of standard bbn see e.g. @xcite ) .
nevertheless , present data still leave room for a non - standard energy density content , or a non standard value for the hubble expansion rate .
for example , the analysis carried out in @xcite shows that a conservative observational range for the helium-4 mass fraction constrains any extra contribution or deficit in the radiation energy content in the range @xmath30 , when parameterizing it in terms of extra effective neutrino species . in view of these considerations ,
the main question we would like to address is whether it is still possible that the bd field @xmath31 contributes during bbn in a non negligible way .
the effect of @xmath31 can be easily discerned from eq .
( [ eqzns2 ] ) .
in particular , its dynamics affects the value of the gravitational constant which scales as @xmath32 and which ties the energy - density to the expansion rate .
moreover , the dynamics of @xmath31 provides extra contributions to the hubble parameter .
the usual assumption is to consider a negligible role of the functions @xmath33 ( the `` potential energy term '' ) and @xmath34 ( the `` kinetic energy term '' ) in eq .
( [ eqzns2 ] ) , is somehow imprecise , since their sum does not correspond in general to a covariantly conserved energy density .
nevertheless they are useful quantities to quantify the contribution of the bd field dynamics to the expansion rate and moreover , their sum is covariantly conserved in the limit of very large @xmath35 and negligible energy density @xmath26 of standard matter which we discuss in this section , see e.g. @xcite . ]
so that the usual bbn scenario is only possibly changed by a different value of the effective gravitational constant or , equivalently , of a re - scaled energy density , see e.g. @xcite .
the same bound previously quoted in term of @xmath36 then translates into @xmath37 . using the covariant conservation of the energy for standard matter @xmath53 , eq .
( [ norma ] ) can be further simplified into @xmath54 we are interested in a situation in which the expansion rate of the universe is of the order of the standard value leading to a successful bbn , and yet the @xmath31 contribution is relevant or even dominant . by inspection of ( [ eqzns2 ] )
we see that in order for these conditions to hold , the energy density @xmath26 should be suppressed by a value of @xmath55 .
furthermore , the new terms , which are dependent on the dynamics of @xmath31 , should _ mimic _ a radiation dominated behavior and compensate for the reduced contribution of ordinary radiation to the expansion rate . in order to see whether this scenario is in fact possible at all
, we start by considering the simplest case , in which ordinary radiation and matter contribute negligibly to the kinematics .
of course , this means that we have to check _ a posteriori _ that , during bbn , it is possible to choose values of @xmath31 large enough so that @xmath26 is sufficiently suppressed .
the leading contribution to ( [ eqzns2 ] ) will then be provided by the dynamics of @xmath31 .
field versus @xmath48 for @xmath75 and @xmath76 .
the field bounces back at @xmath62 where @xmath77 . ]
notice that since @xmath35 is very large , in the interval @xmath78 the two solutions can be expressed with an accuracy better than @xmath79 as _
( a)= _ 0 ( 1 a ) .
plugging eq .
( [ phia ] ) into eq .
( [ pott ] ) ( and neglecting @xmath26 ) one has @xmath80 where the last expression holds for large @xmath35 .
let us summarize the results obtained so far . by considering a potential of the form shown in eq .
( [ pot2 ] ) , the expansion rate of the universe due to a bd field behaves the same way as during a radiation dominated regime .
depending on the sign of the power - index in @xmath50 , there are two possible behaviors for the field @xmath31 , increasing or decreasing with cosmic expansion .
the @xmath31 field rolls very smoothly along the potential , and with comparable `` kinetic '' and `` potential '' energy densities @xmath81 and @xmath82 respectively .
since we derived these results in the limit @xmath83 , we must check that this condition is compatible with the solution found .
indeed , it is easily seen that the term @xmath82 in eq.([pot2 ] ) is independent from the value of @xmath61 , while the term @xmath84 can be made arbitrarily small by choosing @xmath61 arbitrarily large .
this implies that provided we choose the initial value for @xmath31 to be sufficiently large , the contribution of ordinary radiation can be neglected , and the expansion of the universe follows @xmath85 despite the fact that it is driven by a bd scalar field .
in the previous section we have found that there are two solutions for which a bd field dominates cosmic dynamics , while keeping a radiation - dominated expansion as in eq .
( [ radia ] ) .
we turn now to the phenomenological requirement of preserving in this scenario the general agreement between bbn predictions and the observed light element yields . of course
, the scenario outlined in the previous section of a pure @xmath31-dominated universe can not describe the late stages of cosmic evolution .
indeed this would be at variance with the fact that today we require @xmath86 .
moreover , with the potential considered in the previous section , the universe would continue to expand as in a radiation dominated regime until very small redshifts .
we do not address these issues further here , but it is clear that a more complicated dynamics after the bbn period is required in order to bring such kind of models in accordance with `` late '' cosmological and astrophysical observables . in the following ,
in particular we will assume that cmb observations are explained as in the standard scenario .
the set of equations governing bbn , besides equations ( [ eqzns1 ] ) and ( [ eqzns2 ] ) , is the following ( see e.g. @xcite ) 1n_b&=&-3h + & = & -3h(+p ) + & = & _ i(x_j ) + l(m_e / t,_e)&=&_j z_jx_j + ( _ t - h|p|_|p| ) f__(|p| , t)&=&i__[f__e , f_|_e , ] .
the first two equations state respectively the conservation of the total baryon number and energy density in a comoving volume .
the third equation is the boltzmann equation which describes the density of each nuclide species , with the @xmath87 being the rates of interaction averaged over the kinetic equilibrium distribution functions .
the fourth equation accounts for the electric charge neutrality of the universe in terms of the electron chemical potential @xmath88 . finally , the last equation is the boltzmann equation for the neutrino species .
it is clear that the field @xmath31 never explicitly enters these equations but does appear in the evolution of the hubble parameter .
therefore , as long as the evolution of @xmath2 in the bd case is indistinguishable from the standard case , adopting initial conditions ( at whatever @xmath89mev ) for the matter and radiation fields analogous to the standard case , the production of the light elements will also be unchanged .
however , even in the scenario described in the previous section there is one relevant departure , since the hubble factor evolves strictly as @xmath90 , while in the standard case this evolution is modified because @xmath91 annihilation changes the number of effective relativistic degrees of freedom .
note that this phenomenon happens just between the two crucial temperatures for bbn ; @xmath92mev at which the neutron / proton ratio freezes , and @xmath93mev when the deuterium bottleneck opens and light element production begins .
if the bd hubble rate matches one of the two conditions , it might fail to match the second one . to test
if such a difference can be accommodated at all , we have run the bbn code described in @xcite by assuming that the dominant contribution to @xmath2 is provided by the bd field , and its behavior is as for pure radiation as discussed in the previous section . to this end
we replace the standard matter / radiation energy density @xmath26 in the hubble law ( and only there ! ) with the following parametrization in terms of a single neutrino energy density @xmath94 where @xmath95 and @xmath96 . in the limit in which the solution previously found holds ,
this useful parametrization is _
exact_. we ask if there is any value at all for the factor @xmath97 which can accommodate at least @xmath42he and @xmath9h abundances .
since we assume that the standard cosmology holds for @xmath98 , we fix the baryon abundance to that deduced from the cmb , eq .
( [ omcmb ] ) .
note that , since there are two conditions to fulfill with only one parameter , the constraint is non - trivial .
we find that , for @xmath99 the conservative constraint on the @xmath42he mass fraction @xmath100 quoted in @xcite is satisfied , while predicting a deuterium fraction @xmath101 @xmath9h / h @xmath102 , comparable with the observed one @xmath9/h@xmath103 @xcite within 2@xmath104 . for @xmath97 in the range of eq .
( [ boundy ] ) we also found @xmath105 @xmath41he / h @xmath106 , which is consistent with present bounds @xcite , and @xmath107 @xmath43li / h @xmath108 , which remains a factor @xmath109 3 larger than the observed one @xcite .
( light grey ) contours in the @xmath121 plane . ]
it should be noticed that to impose the condition that bd behaves as radiation , or that @xmath125 corresponds to the choice of a particular potential . in principle
one should show that this choice is possible . in order to see
if this is indeed sensible , we use ( [ potenza ] ) in equations ( [ eqzns1])([eqzns2 ] ) and find that a solution can be found by taking the time derivative of these two equations .
the outcome is a system of nonlinear 2@xmath126 order differential equations in @xmath48 and @xmath31 .
there is only one independent initial condition
the choice of @xmath127since @xmath128 can be found by using equation ( [ eqzns2 ] ) , and since equation ( [ eqzns1 ] ) becomes a cubic equation in @xmath129 .
this means that the system becomes uniquely determined and the solution does indeed exist and it is unique .
we have revisited the constraints on a brans - dicke scalar field , arising from the requirement of successful nucleosynthesis .
this is an important question in light of the different roles that a brans - dicke field @xmath31 may play in various sectors of cosmology , such as inflation , baryogenesis and dark energy .
indeed , we have found that the bd field , i.e. newton s constant , can have significantly different values from today s one , and at the same time dominate the energy density making the universe behave as if radiation dominates .
the matter content is not changed at all , so that all physical quantities such as freeze - out temperatures and decay / production rates for elements do not differ from standard ones . in other words , we have found that it is possible for the bd field to mimic the effect of radiation on the evolution of the cosmic scale factor . in such a case @xmath31 is the dominant contributor to the hubble parameter during bbn .
we have put constraints on the bd potential using the observed light elements yields and the bbn predictions in such a modified scenario .
this scenario requires a large initial value for @xmath130 so that the standard contribution of matter to the expansion rate is suppressed by the reduced value of the effective newton constant , as well as potential for the bd field @xmath50 that scales as a power law in @xmath31 , w/^2 . here
@xmath131 if we assume that the local result of the cassini mission also holds on cosmological scales .
we do not offer any suggestion how to explain such an high exponent . by numerically solving the set of bbn equations we have found that it is indeed possible to tune the overall scale of the potential @xmath50 , see eq .
( [ resfin ] ) , so as to obtain final yields for both @xmath9h and @xmath42he compatible with the observed data .
of course , one may worry that large values of @xmath31 at the bbn epoch are not compatible with today s value .
in fact , if we considered the same effective potential valid until today , the field @xmath31 would keep on being the dominant component of the universe , leading to a completely different evolution for the universe , e.g. an extrapolation would lead to value for the field @xmath31 incompatible with today s value for the newton gravitational constant . therefore , after bbn , the potential should be chosen such that the field decreases to acceptable values , and the universe follows a standard evolution .
the fact that today s value of @xmath31 imposes an upper bound on the possibility of having very large values for @xmath132 is tantamount to saying that a fine - tuning of the choice of the potential is necessary in order to allow a huge change in the value of the field from bbn up to now .
finally , we have considered the more general scenario where both standard radiation and the bd field contribute to the expansion rate , in the case that the bd dynamics again corresponds to a pure radiation behavior . in this case too , by comparing the theoretical prediction with light nuclei abundances , we have found that , depending on the value of the bd field during bbn and its contribution to @xmath2 , there is a region in this two - parameter space leading to successful nucleosynthesis .
this region shown in figure [ fig : likely ] extends from the high @xmath31 regime already discussed down to the standard scenario with @xmath86 showing that provided the @xmath31 potential is suitably tuned , basically all values of the bd field can be shown to be compatible with the abundances of both @xmath9h and @xmath42he produced during bbn .
an interesting feature of the scenario considered in this paper is that one is basically free to adjust the time - temperature relation , which analytically means rescaling the initial conditions for the thermodynamical quantities of the fluids in the plasma .
we have limited our discussion to the case where @xmath4 around the bbn epoch has the same evolution it would have in a radiation - dominated plasma , keeping the particle densities fixed at the same value of the standard case as our initial conditions .
this is not a fine - tuning required a priori , but an useful working hypothesis which authomatically ensures that the rates for the scattering processes ( including the nuclear reactions ) are unchanged .
however , one may relax this assumption , and thus the allowed region in the model space may enlarge considerably , to the price of major ( and model - dependent ) modifications in the physical processes in the plasma .
we do not discuss further this point in our phenomenological approach , since our purpose in this article was to show that exotic scenarios where the bd field is driving the expansion rate at the bbn epoch are indeed possible .
nonetheless , this feature should be kept in mind when characterizing specific models ( as e.g. quintessence potentials with solutions tracking radiation ) with a cosmic evolution before and after the bbn epoch determined self - consistently within the theory .
g.m . acknowledges full support by syracuse university during his visit to the physics department , syracuse university .
mt and adf are supported in part by the national science foundation , under grant nsf - phy0354990 , by funds from syracuse university and by research corporation .
adf is supported in part by pparc .
ps is supported , in part , by the deutsche forschungsgemeinschaft under grant no .
sfb 375 and by the european union under the ilias project , contract no .
rii3-ct-2004 - 506222 .
d. n. spergel _ et al .
_ , `` wilkinson microwave anisotropy probe ( wmap ) three year results : implications for cosmology , '' astro - ph/0603449
. s. perlmutter _ et al .
_ [ supernova cosmology project collaboration ] , astrophys .
j. * 517 * ( 1999 ) 565 [ arxiv : astro - ph/9812133 ] .
a. g. riess _ et al .
_ [ supernova search team collaboration ] , astron .
j. * 116 * ( 1998 ) 1009 [ arxiv : astro - ph/9805201 ] .
riess et al .
, astrophys .
j. * 607 * ( 2004 ) 665 [ astro - ph/0402512 ] . s. m. carroll , v. duvvuri , m. trodden and m. s. turner , phys .
d * 70 * ( 2004 ) 043528 [ arxiv : astro - ph/0306438 ] .
s. m. carroll , a. de felice , v. duvvuri , d. a. easson , m. trodden and m. s. turner , phys .
d * 71 * ( 2005 ) 063513 [ arxiv : astro - ph/0410031 ] .
s. capozziello , s. carloni and a. troisi , arxiv : astro - ph/0303041 . c. deffayet , g. r. dvali and g. gabadadze , phys .
d * 65 * ( 2002 ) 044023 [ arxiv : astro - ph/0105068 ] .
k. freese and m. lewis , cardassian expansion : a model in which the universe is flat , matter phys .
b * 540 * ( 2002 ) 1 [ arxiv : astro - ph/0201229 ] .
n. arkani - hamed , s. dimopoulos , g. dvali and g. gabadadze , arxiv : hep - th/0209227 . g. dvali and m. s. turner , arxiv : astro - ph/0301510 . n. arkani - hamed , h. c. cheng , m. a. luty and s. mukohyama , jhep * 0405 * ( 2004 ) 074 [ arxiv : hep - th/0312099 ] . c. wetterich , nucl .
b * 302 * ( 1988 ) 668 .
b. ratra and p. j. e. peebles , phys .
d * 37 * ( 1988 ) 3406 .
r. r. caldwell , r. dave and p. j. steinhardt , phys .
* 80 * ( 1998 ) 1582 [ arxiv : astro - ph/9708069 ] .
l. m. wang , r. r. caldwell , j. p. ostriker and p. j. steinhardt , astrophys .
j. * 530 * ( 2000 ) 17 [ arxiv : astro - ph/9901388 ] .
m. sahlen , a. r. liddle and d. parkinson , phys .
d * 72 * , 083511 ( 2005 ) [ arxiv : astro - ph/0506696 ] .
y. fujii and k. maeda , _ the scalar - tensor theory of gravitation _
, cambridge university press 2003 . c. brans and r. h. dicke , phys .
* 124 * ( 1961 ) 925 .
r. h. dicke , phys .
* 125 * ( 1962 ) 2163 .
s. m. carroll , a. de felice and m. trodden , phys .
d * 71 * ( 2005 ) 023525 [ arxiv : astro - ph/0408081 ] .
l. perivolaropoulos , phys .
d * 67 * ( 2003 ) 123516 [ arxiv : hep - ph/0301237 ] . v. faraoni , int .
j. mod .
d * 11 * ( 2002 ) 471 [ arxiv : astro - ph/0110067 ] .
e. gunzig , a. saa , l. brenig , v. faraoni , t. m. rocha filho and a. figueiredo , phys .
d * 63 * ( 2001 ) 067301 [ arxiv : gr - qc/0012085 ] .
a. riazuelo and j. p. uzan , phys .
d * 66 * ( 2002 ) 023525 [ arxiv : astro - ph/0107386 ] .
j. p. uzan , phys .
d * 59 * ( 1999 ) 123510 [ arxiv : gr - qc/9903004 ] .
f. perrotta and c. baccigalupi , nucl .
. suppl .
* 124 * ( 2003 ) 72 [ arxiv : astro - ph/0205245 ] . f. perrotta and c. baccigalupi , phys .
d * 65 * ( 2002 ) 123505 [ arxiv : astro - ph/0201335 ] .
f. perrotta , c. baccigalupi and s. matarrese , phys .
d * 61 * ( 2000 ) 023507 [ arxiv : astro - ph/9906066 ] .
v. pettorino , c. baccigalupi and g. mangano , jcap * 0501 * ( 2005 ) 014 [ arxiv : astro - ph/0412334 ]
. v. pettorino , c. baccigalupi and f. perrotta , jcap * 0512 * , 003 ( 2005 ) [ arxiv : astro - ph/0508586 ] .
t. chiba , phys .
d * 64 * ( 2001 ) 103503 [ arxiv : astro - ph/0106550 ] .
x. l. chen and m. kamionkowski , phys .
d * 60 * ( 1999 ) 104036 [ arxiv : astro - ph/9905368 ] .
d. f. torres , phys .
d * 66 * ( 2002 ) 043522 [ arxiv : astro - ph/0204504 ] . c. santos and r. gregory , annals phys .
* 258 * ( 1997 ) 111 [ arxiv : gr - qc/9611065 ] .
a. de felice and m. trodden , phys .
d * 72 * ( 2005 ) 043512 [ arxiv : hep - ph/0412020 ] . c. y. chen , y. g. shen and b. feng , arxiv : hep - ph/0508059 .
b. feng , h. li , m. z. li and x. m. zhang , phys .
b * 620 * ( 2005 ) 27 [ arxiv : hep - ph/0406269 ] .
h. li , m. z. li and x. m. zhang , phys .
d * 70 * ( 2004 ) 047302 [ arxiv : hep - ph/0403281 ] . c. mathiazhagan and v. b. johri , class .
* 1 * ( 1984 ) l29
. j. d. barrow and k. i. maeda , nucl .
b * 341 * ( 1990 ) 294 .
j. steinhardt and f. s. accetta , phys .
lett . * 64*(1990 ) 2740 .
a. m. green and a. r. liddle , phys .
d * 54 * ( 1996 ) 2557 [ arxiv : astro - ph/9604001 ] .
a. r. liddle and d. wands , phys .
d * 45 * ( 1992 ) 2665 .
r. crittenden and p. j. steinhardt , graceful exit in extended inflation and implications for density phys .
b * 293 * ( 1992 ) 32 [ arxiv : astro - ph/9207002 ] .
d. la and p. j. steinhardt , phys .
lett . *
62 * , 376 ( 1989 ) [ erratum - ibid .
* 62 * , 1066 ( 1989 ) ] .
a. l. berkin , k. i. maeda and j. yokoyama , phys .
* 65 * , 141 ( 1990 ) .
s. m. carroll and m. kaplinghat , phys .
d * 65 * ( 2002 ) 063507 [ arxiv : astro - ph/0108002 ] .
a. coc , k. a. olive , j. p. uzan and e. vangioni , phys .
d * 73 * , 083525 ( 2006 ) [ astro - ph/0601299 ] .
b. bertotti , l. iess and p. tortora , nature * 425 * ( 2003 ) 374 . t. clifton , d. f. mota and j. d. barrow , mon . not .
* 358 * ( 2005 ) 601 [ arxiv : gr - qc/0406001 ] . s. eidelman _ et al . _
[ particle data group ] , phys .
b * 592 * ( 2004 ) 1 .
d. kirkman , d. tytler , n. suzuki , j. m. omeara and d. lubin , astrophys .
j. suppl .
* 149 * ( 2003 ) 1 [ arxiv : astro - ph/0302006 ] .
a. cuoco , f. iocco , g. mangano , g. miele , o. pisanti and p. d. serpico , int .
j. mod .
phys . a * 19 * ( 2004 ) 4431 [ arxiv : astro - ph/0307213 ] .
d. serpico , s. esposito , f. iocco , g. mangano , g. miele and o. pisanti , jcap * 0412 * ( 2004 ) 010 [ arxiv : astro - ph/0408076 ] .
r. h. cyburt , phys .
d * 70 * ( 2004 ) 023505 [ arxiv : astro - ph/0401091 ] .
t. damour and c. gundlach , phys .
d * 43 * ( 1991 ) 3873 .
k. a. olive and e. d. skillman , astrophys .
j. * 617 * ( 2004 ) 29 [ arxiv : astro - ph/0405588 ] .
t.m.bania , r.t .
rood , and d.s .
balser , nature * 415 * ( 2002 ) 54 . | we reconsider constraints on brans - dicke theories arising from the requirement of successful big bang nucleosynthesis .
such constraints typically arise by imposing that the universe be radiation - dominated at early times , and therefore restricting the contribution that a brans - dicke scalar could make to the energy budget of the universe .
however , in this paper we show how the dynamics of the brans - dicke scalar itself can mimic a radiation - dominated kinematics , thereby allowing successful nucleosynthesis with a sizable contribution to the total cosmic energy density .
in other words newton s constant may dynamically acquire values quite different from that today , even though the evolution mimics radiation domination .
this possibility significantly relaxes the existing bounds on brans - dicke fields , and opens the door to new possibilities for early universe cosmology .
the necessary fine tunings required by such an arrangement are identified and discussed .
+ _ dedicated to rafael sorkin , on the occasion of his 60th birthday , + and to celebrate his wonderful contributions to physics _ |
synthetic fingerprint generation , minutiae distribution , transportation problem , earth mover s distance , fingerprint reconstruction , identification test , fingerprint individuality .
_ at the fifteenth international conference on pattern recognition held in september 2000 , about 90 participants , most of them with some background in fingerprint analysis ( quoted from page 293 in @xcite ) were shown four fingerpint images ( reproduced in figure [ figsurvey ] ) .
they were told that three images were acquired from real fingers and one of these prints was synthetically generated and they were asked to identify the artificial print .
only 23% of the participants correctly chose the synthetic print ( image ( a ) in figure [ figsurvey ] ) . _
_ in this paper , we present a method that is able to perform the task at which the human experts failed : we introduce a test of realness which can separate real from synthetic fingerprint images with very high accuracy .
we applied the method ( a detailed description follows in section [ histograms : scn ] and [ discrimination : scn ] ) to these four images and the proposed method distinctly identifies the correct image ( computing the difference of minutiae histogram distances explained in section [ emddiff ] yields a negative score for image ( a ) in figure [ figsurvey ] and positive scores for images ( b - d ) ) . _
today , many commercial , governmental and forensic applications rely on fingerprint recognition for verifying or establishing the identity of a person . among these , methods building on minutiae matching play an eminent role . usually , matching routines are not only tested on real fingers , but in order to provide for theoretically unlimited sample sizes , synthetic fingerprint generation systems such as sfinge by cappelli _
( @xcite ) have been developed in the past .
independently , methods have been developed to reconstruct fingerprints from minutiae templates ( cf .
both methodologies are very relevant in many application areas : * constructing synthetic fingerprint images facilitates the cheap creation of very large databases for testing and comparing the performance of algorithms in verification and identification scenarios . *
ground truth data is provided for evaluating the performance of forensic experts @xcite as well as minutiae extraction algorithms . * matching performance of low - quality and latent fingerprints
is improved @xcite .
* fingerprint reconstruction can be a building block for solving interoperability problems , e.g. on comparing fingerprints acquired from different sensors @xcite .
* research in this area raises the awareness for aspects of security , privacy and data protection bearing in mind that an attacker may utilize existing techniques for creating a spoof and prepare a presentation attack . * mixing prints of two or more real fingers has been proposed for generating virtual identities , obscuring private information or creating cancelable templates @xcite . of course , synthetic prints should be as real as possible pertaining to all properties and features which are relevant for fingerprint recognition , especially with respect to their minutiae distribution .
otherwise , a human may be fooled by the look of a synthetic print ( see the motivational example given before the introduction taken from page 293 in @xcite ) , but their eligibility e.g. for evaluating fingerprint recognition algorithms may be challenged and results obtained on artificial databases would be insignificant . [ [ a - unifying - concept - of - the - correct - minutiae - distribution . ] ] a unifying concept of the correct minutiae distribution . + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + fingerprint synthesis and fingerprint reconstruction have been treated for a long time as different tasks .
this can be well conceived on the background that the issue of realistic minutiae distributions has only played a subordinate role in theoretical model building and practical research . in our contribution
we provide for a simple method to assess minutiae distributions of single fingerprints as well as of samples of fingerprints .
we demonstrate that after training , this allows to decide that minutiae patterns of synthetically generated fingerprints are not correct. in particular , we believe that including realistic minutiae distributions leads to a unified concept in which synthetic fingerprint generation and fingerprint reconstruction are in fact two sides of the same coin .
we will return to this point in section [ labeltwosidesofthesamecoin ] .
[ [ fingerprints - of - fingerprints . ] ] fingerprints of fingerprints .
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + minutiae histograms ( mhs ) introduced below assign any given fingerprint image a fixed length feature vector .
this feature vector is not only highly potent to discriminate real fingerprints from fingerprints synthetically generated by the only publicly available current state of the art system , in a preliminary study we additionally demonstrate that this new feature vector is also highly discriminatory among real fingerprints as well .
given reliably extracted minutiae and sufficent overlap of latent fingerprint images , mhs allow for fast and effective matching .
more precisely , it seems that already the second order mhs reflecting minutiae pairs only , promise a high potential awaiting to be yet unleashed .
note that higher order mhs can be assessed via minutiae triplets , quadruplets etc .
while local minutiae statistics are widely used for matching , our contribution is to consider global minutiae statistics and measure distances by a metric that effectively captures human perception : the earth mover s distance .
as made clear above , in our contribution we focus on the role of minutiae distributions .
this issue is currently gaining momentum in the scientific community , cf .
in contrast to @xcite , who use parametric mixture models , we take a nonparametric statistical approach .
we begin our exposition with biological principles currently believed to govern minutiae formation and their distribution . to date ,
however , these principles are still not satisfactorily understood @xcite .
subsequently , we put the sfinge method into context with these biological hypotheses and describe an alternate method based on real fingerprint images for generating latent fingerprints , contrasting in this and other aspects .
thereafter , we allude to the similarities between fingerprint generation and reconstruction which will resurface when we propose improvements in synthetic fingerprint generation .
we conclude this section by laying out the plan of the paper .
[ [ fingerprint - formation - guided - by - merkel - cells . ] ] fingerprint formation guided by merkel cells .
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + kcken and champod propose a model for fingerprint formation @xcite that has two major influence factors : growth forces which create mechanical compressive stress @xcite on the one hand , and merkel cells rearranging from a random initial configuration into lines minimizing the compressive stress and inducing primary ridges on the other hand .
merkel cells interact with each other in reaction - diffusion systems of short range attraction and long range repulsion @xcite . based on empirical evidence from embryonic volar tissue evolution ( e.g. @xcite ) , kcken and champod
let solutions of suitable partial differential equations propagate from three centers : one along the flexion crease , one at the volar pad ( the core area ) and one from the nail furrow . based on specific parameter choices
, this process eventually forms a ridge line pattern featuring minutiae .
kcken and champod analyzed the images resulting from simulation runs of their growth model and discovered qualitative differences in comparison to real fingerprints .
they conclude that with respect to the natural variability of arrangements of minutiae only an empirical acquisition of genuine fingerprints will provide an adequate source of data @xcite .
[ [ sfinge . ] ] sfinge .
+ + + + + + + in a nutshell , growing fingerprint patches containing no minutiae are generated starting from a number of randomly located points by the iterative application of gabor filters according to a previously generated global orientation field @xcite and a non - constant ridge frequency pattern .
whenever patches meet , minutiae are produced whenever necessary for the consistency of the global ridge pattern . a detailed description of the sfinge method by cappelli _
_ can be found in @xcite , chapter 6 in @xcite and chapter 18 in @xcite . in fact ,
the sfinge model silently assumes biological hypotheses of fingerprint pattern formation that are slightly different from the ones described above .
first , fingerprint patterns no longer propagate from a well defined system of three original sources but rather from a multitude of sources at random locations .
secondly , a main governing principle for minutiae creation lies in the compatibility of ridge patterns whenever growing patches touch .
these hypotheses in itself are very interesting and can be viewed as intuitively natural ; and their validity can be assessed with our methodology . our test of realness ( section [ sectestrealness ] ) yields ,
however , that these hypotheses explain the process of minutiae formation not satisfactorily ( section [ secresults ] ) .
notably , this shortfall of sfinge can not be explained by imprint distortions ( which are already included in the sfinge model ) and the lack of a true 3d model , as both would account mostly for neighboring bin perturbations or none at all .
using the earth mover s distance @xcite , precisely distinguishes between these minor perturbations and the major deviations we found ( for a detailed discussion cf .
@xcite ) .
[ [ further - methods . ] ] further methods .
+ + + + + + + + + + + + + + + + further methods for generating synthetic fingerprint images were proposed in 2002 by araque et al .
@xcite and in 2003 by bicz @xcite .
the methods of araque et al .
and sfinge are similar in spirit and they both rely on the global orientation field model of vizcaya and gerhardt @xcite for creating an orientation field .
examples of synthetic fingerprints generated by this method are shown in figure [ figsynthetic ] ( a ) and ( b ) .
viewing fingerprints as holograms ( cf .
@xcite ) , a given global orientation field yields the continuous phase onto which so called spiral phases can be added giving minutiae at specific predefined locations while other minutiae occur due to continuity constraints .
this allows for synthetic fingerprint generation and four example images generated by the software of bicz @xcite are displayed in figure [ figsynthetic ] ( c ) to ( f ) .
results reported in section [ secresults ] show that the method proposed in this paper can also discriminate between synthetic fingerprints according to bicz and real fingerprint images . + + [ [ creating - forensic - fingermarks - from - real - prints . ] ] creating forensic fingermarks from real prints .
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + the approach by rodriguez
_ et al . _
@xcite , since the minutiae are based on real fingers , does not suffer from unrealistic minutiae configurations .
it is however subject to increased time , money and data protection constraints .
forensic fingermarks ( latent fingerprint images ) are simulated in a semi - automatic way : fingers of volunteers are recorded using a livescan device . during a period of about 30 seconds , each person performs a series of predefined movements which results in fingerprint images with various distortions .
images are captured at a rate of four frames per second .
minutiae are automatically extracted followed by manual inspection and correction of falsely extracted or missed minutiae . starting with this ground truth data set
, latents are simulated using a region containing a cluster of 5 - 12 minutiae from the real fingerprint images .
one use case for these simulated latents is to test the minutiae marking performance of human experts which can be evaluated against the avaivable ground truth information .
another application area is benchmarking the identification performance of afis software . [
[ reconstruction . ] ] reconstruction .
+ + + + + + + + + + + + + + + various researches have shown that fingerprints can be reconstructed from minutiae templates @xcite .
there are two major approaches for the automatic reconstruction of fingerprint images : first , the iterative application of gabor filters as in sfinge , and second , the usage of amplitude- and frequency - modulated ( am - fm ) functions @xcite . in both cases ,
the first step is the estimation of an orientation field which fits the minutiae pattern .
fingerprint reconstruction from minutiae templates and the generation of synthetic images follow similar principles .
however , the goal in the reconstruction scenario is to produce the same number of minutiae at the same locations and with same direction and type as in the template . [ [ synthetics - vs .- spoofs . ] ] synthetics vs. spoofs .
+ + + + + + + + + + + + + + + + + + + + + + the mh based method proposed in this paper does not perform liveness detection @xcite . in fact
, we would like to stress that if an attacker can produce a high quality spoof containing the same minutiae as the imitated alive finger and puts this spoof finger on fingerprint sensor , the proposed method will classify the acquired fingerprint image as originating from a real finger with respect to the minutiae distribution . only attacks with spoofs based on synthetic images and synthetic minutiae distributions are addressed in this work . in the next section ,
we introduce second order extended minutiae histograms ( mhs ) which for this paper we just call minutiae histograms : statistics of pairs of two minutiae are used in combination with minutiae type information and interridge distances to obtain a fingerprint of fingerprints . in section [ discrimination :
scn ] , extended mhs are applied for classifying a fingerprint into one of the two categories , real or synthetic .
tests on the 12 publicly available databases of fvc2000 , fvc2002 and fvc2004 show the discriminative power of this approach . in section [ secimprovement
] , detailed suggestions for the generation of more realistic synthetic fingerprints are given . in section [ secidentification ] , the suitability of minutiae histograms for identification purposes is investigated and in section [ secindividuality ] , mhs are proposed for the quantification of fingerprint evidence .
section [ discussion : scn ] conludes with a discussion and states topics of further research . + [
cols="^,^,^",options="header " , ] [ secresults ] & & & & + & set ii & set iii & set ii & set iii & set ii & set iii & set ii & set iii + fvc2000 & 93.3 & 90.0 & 95.0 & 83.8 & 98.3 & 95.0 & 95.5 & 89.6 + fvc2002 & 86.7 & 88.8 & 90.0 & 86.3 & 85.0 & 78.8 & 87.2 & 84.6 + fvc2004 & 81.7 & 55.0 & 81.7 & 67.5 & 90.0 &
70.0 & 84.5 & 64.2 + + & set ii & set iii + & 83.3 & 92.5 + + & & & & + fvc2000 & 100.0 & 92.5 & 100.0 & 87.5 & 100.0 & 97.5 & 100.0 & 92.5 + fvc2002 & 100.0 & 97.5 & 100.0 & 95.0 & 100.0 & 90.0 & 100.0 & 94.2 + fvc2004 & 95.0 & 72.5 & 90.0 & 83.8 & 98.3 & 97.5 & 94.4 & 83.6 + the results on all available fvc databases show that the proposed method by extended 2d - mhs is able to separate real from synthetic prints with very high accuracy . on the training sets of all databases of fvc2000 and fvc2002 ,
the classification performance of the combined feature set was 100% , and for the corresponding test sets , the performance was in the range from 87.5 to 97.5% . on the whole ,
the image quality in the databases of fvc2004 is clearly lower compared to the quality of images in previous competitions .
hence , it it is more challenging to avoid errors during the automatic extraction of minutiae from theses images .
we used a commercial off - the - shelf software for minutiae extraction .
the good discriminative power of 2d - mhs alone is not surprising upon inspection of a 2d visualization by multidimensional scaling ( e.g. ( * ? ? ?
* chapter 14 ) ) of the mutual 2d - mh distances from table [ tabavgemd ] in figure [ minutiaeangledist - hist - wassersteinmds : fig ] . in the left display ,
mhs from real fingers cluster in the middle of the left side while mhs from synthetic fingerprints come to lie in the upper right ( for the year 2000 ) and closer to the center towards the bottom ( for the years 2002 and 2004 ) .
in fact , we can see that the algorithms leading to minutiae formation have obviously undergone changes between the years 2000 and 2004 , however , only moving the mhs moderately closer to realness. in fvc2002 , sfinge version 2.51 was used and in fvc2004 sfinge version 3.0 .
additionally , we tested the performance of 2d - mhs for discriminating between artificial fingerprint images generated by the software of bicz @xcite and real fingerprint images . to this end , we generated 110 synthetic fingerprints which were grouped into three sets as listed in table [ tabsets ] , and compared them against the 110 images of fvc2000 database 1 . on set
ii , a classification performance of @xmath0 was achieved and on set iii , @xmath1 of the images were correctly classified .
recall from section [ secconstructionandreconstruction ] that the sfinge model silently extends the well - standing biological hypothesis of fingerprint pattern formation due to three converging ridge systems ( for a brief discussion , c.f .
@xcite ) , to a multitude a converging ridge systems starting at random locations ; this process is the governing principle for minutiae formation .
our work shows that this remarkable hypothesis can be tested and the tests show that the minutiae pattern formation appears to be of a more complex nature .
it would be of interest to test fingerprint images generated by @xcite , @xcite and other reseachers using extended minutiae histograms , once they become available .
one possibility to improve sfinge is to modify the fingerprint generation process in such a way that the resulting fingerprint images have the following properties : * mhs of synthetic prints should follow the distribution of mhs estimated from a database of real fingers . *
more subtly also considering minutiae types , the relation of endings to bifurcations should resemble the relation and its distribution observed in real fingers .
* similarly , the distribution of interridge distances should be similar to those of real fingers acquired at the same resolution .
[ [ labeltwosidesofthesamecoin ] ] two sides of the same coin + + + + + + + + + + + + + + + + + + + + + + + + + + an alternative option for the creation of synthetic fingerprints that pass the proposed test of realness is to consider synthetic fingerprint generation as a reconstruction task .
traditionally , the construction of artificial fingerprints is associated with an unknown outcome regarding the number , location , direction and type of minutiae , whereas reconstruction aims at the generation of a fingerprint image which has best possible similarity in terms of minutiae properties to a given template .
basically , we propose to focus on the generation of a realistic minutiae template and `` all these other things above shall be added unto '' by the existing reconstruction methods . here is a possible outline of this approach : * first , a feasible foreground and orientation field is constructed , e.g. using a global model like @xcite which is able to incorporate the empirical distribution of observed pattern types ( henry - galton classes ) and singular points @xcite . *
secondly , a realistic number of minutiae @xmath2 is drawn from the empirical data of minutiae in fingerprints with the previously determined foreground size and pattern type .
* thirdly , an inital minutiae template is obtained by choosing @xmath2 points e.g. randomly on the foreground as minutiae locations and for each location , the minutiae direction is set to the local orientation @xmath3 or @xmath4 by chance . *
fourthly , the 2d - mh of the minutiae template is computed and it is modified iteratively until it passes the test of realness , i.e. the emd between the current template and the average 2d - mh of real fingerprints is below an acceptable threshold .
modification operations are the deletion and addition of minutiae and flipping of the minutiae direction by 180@xmath5 .
the implementation for the computation of the emd allows to analyze the flow of mass , so that the bins in the 2d - mh can be indentified which contribute above the ordinary to the total costs , and thus , the minutiae pairs that are the most unlikely in comparison to the empirical distribution . *
fifthly , minutiae types ( ending and bifurcation ) are assigned , based on the empirical distribution in real finger patterns . * finally , the fingerprint image is reconstructed using e.g. gabor filters or the am - fm model .
the interridge distances are checked for deviations from the empirical data obtained from real fingerprints and if required , the interridge distance image is adjusted and the reconstruction step is repeated . *
optionally , noise can be simulated for copies of the constructed image , and if desired , they can be rotated , translated and nonlinearly distorted .
, title="fig:",scaledwidth=45.0% ] , title="fig:",scaledwidth=45.0% ] up to this point , we applied minutiae statistics for classifying a fingerprint as real or synthetic . in this section
, we explore the potential of mhs for identifying individuals .
the approach described previously based on mhs allows to represent a fingerprint as a fixed - length feature vector . obtaining a fixed - length feature vector representation marks the grand goal for biometric template protection schemes and
has previously been achieved by xu _
_ in their spectral minutiae representation @xcite .
robust fixed - length feature vectors also appear highly promising for an identification scenario in which a large database is searched for a query fingerprint . up to this point ,
the minutiae distributions represented by mhs were compared using the earth mover s distance . for the purpose of identification ,
we compute the histogram intersection to better handle partial overlap @xcite .
this approach is related to `` geometric hashing '' @xcite : a hash table is built by quantizing geometric objects like minutiae triplets which were used in @xcite . in a first test on fvc2000 db2 , we achieved average access rates ( average part of the database that is accessed using an incremental search strategy until the corresponding finger is found ) between 2.33% and 4.01% using unnormalized 4d - mhs with 20 bins for differences between minutiae directions , 20 bins for euclidean distances between minutiae locations , 20 bins for the angle of the relative location of the second minutiae with respect to the first and 4 bins for minutiae type combinations .
we chose this database , because it was used in tests measuring the indexing performance by other researchers : de boer _ et al .
_ tested three different features ( orientation field , finger code , minutiae triplets ) and their combination on this database and they reported rates between 1.34% and 7.27% @xcite .
proposed minutiae cylinder code and locality - sensitive hashing for fingerprint indexing @xcite and report an average access rate of 1.72% for this database .
as we used , in contrast to these two studies , an `` off the shelf '' minutiae extractor ( ) , and as the performance hinges on correct minutiae extraction ( see figure [ figindex ] and the discussion below ) we deem our results very promising upon more `` correct '' minutiae extraction . in our test , we used the sum of intersections between corresponding bins as a score ( bis = bin intersection score ) and sorted the list of fingers in the database in descending order . in 597 out of 770 searches ( 77 @xmath6 ) ,
the correct finger was ranked first in the list . if the search is narrowed down to templates with 30 or more minutiae , than it ranked first in 393 out of 441 searches ( 89 @xmath6 ) .
we inspected the cases in which a larger portion of the database were accessed .
it turned out that the main reason lies in minutiae extraction errors ( missing and spurious minutiae ) which , of course , have a negative impact on the score : missing minutiae reduce the intersection between minutiae histograms of templates from the same finger , spurious minutiae can increase the intersection between histograms from templates of different fingers .
a typical example is shown in figure [ figindex ] .
another reason lies in almost no overlap areas between the two prints compared .
we stress at this point that our bis has been designed to alleviate small overlaps .
it is obvious that for a fair comparison between different minutiae - based indexing methods , the same minutiae templates have to be used . in doing so
, the influence of different minutiae extractors on the identification performance can be eliminated and only then results become comparable .
it is of interest to quantify the impact of different minutiae extractors and different fingerprint image enhancement techniques on the identification performance , but these comparisons are beyond the scope of this study .
this first test shows the suitability of mhs for identification purposes and this direction deserves further research .
universality , collectability , permanence and uniqueness are properties which render an anatomical or behavioral trait useful for biometric recognition @xcite .
permanence of the fingerprint pattern was scrutinized by galton @xcite more than a century ago and it was later confirmed that the pattern s development is finalized at an estimated gestational age of 24 weeks @xcite .
uniqueness of fingerprints is commonly assumed by all researchers and practitioners dealing with fingerprint recognition .
fingerprint individuality has never been proven , but there is a long history of models attempting to explain and quantify fingerprint individuality starting with galton in 1892 @xcite to the present day .
stoney gives an overview over 10 major models formulated between 1892 and 1999 in chapter 9 of @xcite .
for recent additions , please see @xcite and the references therein .
there are two broad categories of models : first , mathematical models trying to encompass the distribution of features extracted from observed prints . and secondly , biology based models about randomness during the formation of friction ridge skin in prenatal development of human life @xcite . notwithstanding the lack of proof of uniqueness
, fingerprints have a long success story in commercial , governmental and forensic applications @xcite . as a consequence of an on - going reformation process @xcite
, in the future forensic experts may have to quantify the weight of evidence with probabilities and errors rates instead of a binary decision @xcite . by their very nature ,
mhs ( 2d- , 4d- , second and higher order ) as empirical realisations of an underlying to date unknown
true minutiae distribution precisely yield such probabilities .
for instance for a given confidence level @xmath7 , for each true finger @xmath8 multiply represented in a database , a neighborhood @xmath9 in the space of corresponding ( possibly higher order and/or extended ) mhs can be computed , via bootstrapping , say , such that for a query fingerprint @xmath10 with mh @xmath11 under the null - hypothesis that @xmath10 is another imprint of @xmath8 , @xmath12 where @xmath13 is the empirical probability .
the use of the corresponding test , of course , relies on prior thorough statistical studies of these true distributions .
further research , exploiting the abundance of optional statistical methods , may choose likelihood ratios @xcite , say , for improved quantification .
a crucial point is that the underlying minutiae statistics have to be based on a large number of real fingers .
minutiae should be manually marked or in semi - automatic fashion , a human should inspect automatically extracted minutiae in order to avoid minutiae extraction errors and their influence on the mhs . ideally , interridge distances or related measures like e.g. ridge count
would be taken into account and additional information would be available for minutiae templates , including age , body height , sex and ethnicity @xcite . this would enable the computation of more sophisticated statistics , e.g. the probability that a print with certain minutiae histogram stems from a person of certain age group . in summary , mhs can become a useful tool for forensic experts who are requested to quantify the weight of fingerprint evidence in court based on empirical ground truth data .
the proposed mhs capture and comprise relevant information of fingerprints , namely the full distribution of minutiae , which enable to separate current state - of - the - art synthetic fingerprints from prints of real fingers .
this study reveals a fundamental difference between natural finger pattern formation and state - of - the - art synthetic fingerprint generation processes . as a consequence
, any results obtained on existing databases of synthetic fingerprints should be regarded with caution and may not reflect the performance of a fingerprint comparison software in a real - life scenario .
a performance evalution of fvc2004 showed that the behavior of the algorithms over db4 ( the database consisting of synthetic prints ) was , in general , comparable to that on the real databases @xcite .
however , this can also be interpreted as an indicator that the participating algorithms are not yet optimized for the specifics of the empirical distribution of minutiae in real fingerprints .
another direction of research beyond the ones discussed in sections [ secimprovement ] to [ secindividuality ] can go towards the identification of additional features for the discrimination between real and synthetic prints . in the context of presentation attack detection
, it would be highly desirable to detect a reconstructed fingerprint , even if this property can not be infered from the minutiae distribution .
we would also like to explore the suitability of minutiae histograms for security applications , e.g. in combination with the fuzzy commitment scheme or the fuzzy vault scheme , and for the generation of cancelable templates . in this research
, we used a second order minutiae statistic reflecting the covariance structure of the minutiae distributions which turned out to be highly discriminatory for our task . exploring the discriminative potential of higher order minutiae histograms is beyond the scope of this paper and
remains an interesting endeavor .
the authors thank raffaele cappelli and davide maltoni for sharing the images shown in figure [ figsurvey ] , pedro vizcaya for sharing the images displayed in figure [ figsynthetic ] ( a ) and ( b ) , and wieslaw bicz for sharing the software to generate the images in figure [ figsynthetic ] ( c ) to ( f ) .
the authors gratefully acknowledge the support of the felix - bernstein - institute for mathematical statistics in the biosciences and the volkswagen foundation . c. m. rodriguez , a. de jongh , and d. meuwly . introducing a semi - automatic method to simulate large numbers of forensic fingermarks for research on fingerprint identification . ,
57(2):334342 , march 2012 .
q. zhao , a. k. jain , n.g .
paulter , and m. taylor .
fingerprint image synthesis based on statistical feature models . in _ proc .
5th conf . on biometrics : theory , applications and systems ( btas ) _ , pages 2330 , washington , d.c .
, september 2012 . c. neumann , c. champod , r. puch - solis , n. egli , a. anthonioz , and a. bromage - griffiths .
computation of likelihood ratios in fingerprint identification for configurations of any number of minutiae .
, 52(1):5464 , january 2007 .
e. gutirrez - redomero , n. rivaldera , c. alonso - rodrguez , l. m. martn , j. e. dipierri , m. a. fernndez - peire , and r. morillo . are there population differences in minutiae frequencies ? a comparative study of two argentinian population samples and one spanish sample . ,
222(1 - 3):266276 , october 2012 . | in this study we show that by the current state - of - the - art synthetically generated fingerprints can easily be discriminated from real fingerprints .
we propose a non - parametric distribution based method using second order extended minutiae histograms ( mhs ) which can distinguish between real and synthetic prints with very high accuracy .
mhs provide a fixed - length feature vector for a fingerprint which are invariant under rotation and translation .
this test of realness can be applied to synthetic fingerprints produced by any method . in this work , tests are conducted on the 12 publicly available databases of fvc2000 , fvc2002 and fvc2004 which are well established benchmarks for evaluating the performance of fingerprint recognition algorithms ; 3 of these 12 databases consist of artificial fingerprints generated by the sfinge software .
additionally , we evaluate the discriminative performance on a database of synthetic fingerprints generated by the software of bicz versus real fingerprint images .
we conclude with suggestions for the improvement of synthetic fingerprint generation . |
our turbulent active sun possesses a rather regular magnetic cycle that has been observed directly through the photospheric emergence of sunspots for about four centuries now .
this magnetic flux emergence , which occurs daily at the sun s surface , appears not only at large scales ( sunspots and active regions ) but also at scales as small as the intra - network field and there is indication that even smaller scale flux emerges at the solar surface , which remains unresolved with the present magnetographs @xcite .
however , since a huge amount of flux is brought toward the photosphere by solar active regions and because of the well - established robust properties of such large - scale emergence , we focus our studies on the understanding of the dynamical evolution of large and strong magnetic structures rising from the solar interior to the surface . the appearance of active regions can be summarized in a few steps : small magnetic bipoles first emerge all over the future active region and then merge to form sunspots at the outer edges of the emerging flux region ( efr ) .
the growing phase of a typical efr lasts about 3 - 5 days .
the flux in an active region is usually larger than @xmath1 in each polarity .
the orientation of the bipole may be arbitrary at first , but generally in 1 - 3 days it becomes slightly tilted ( @xmath2-@xmath3 ) with respect to the east - west direction , with the leading polarity closer to the equator , thus conforming to joy s law @xcite .
the tilt angles of individual active regions show a large scatter about the mean but the measurements indicate a tendency for the mean value to linearly increase with the latitude of emergence @xcite .
moreover , various asymmetries have been detected in efrs in particular in the morphology of the leading and trailing spots .
indeed , the leading polarity generally appears more intense and concentrated than the trailing one which tends to be more fragmented .
another interesting feature of efrs is their inherent twist . indeed ,
careful multi - wavelength analysis of flux emergence in active regions ( e.g. * ? ? ?
* ) indicates that they appear with a certain amount of twist ( i.e. , a non - zero current along their axis ) .
this twist is then thought to be responsible for the characteristic pattern known as tongues observed in newly emerging regions @xcite and identified as a major ingredient in mass ejections or flares at the solar photosphere ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
other structures such as an annular shape of the magnetic field around active regions have been observed through various instruments ( e.g. , * ? ? ?
* ) but never really reported as identified structures . however , as we will see , they can serve as strong constraints on models and simulations and the relevant physical processes responsible for their existence will be pointed out .
with all those observational constraints in mind , theoreticians and modelers have tried to analyze in detail the dynamical evolution of strong magnetic structures from the base of the convection zone ( cz ) where they are created to the surface where they appear as sunspots .
a possible scenario for such an evolution would be that strong toroidal fields are created at the base of the cz by the differential rotation shearing a preexisting poloidal field in the tachocline ( see landmark papers of @xcite , recent simulations of @xcite and a review on the solar dynamo theory by @xcite for example ) .
after a certain period of storage in the subadiabatic layer below the cz @xcite , this toroidal field becomes subject to magnetic buoyancy instabilities @xcite and organizes itself in buoyant arched magnetic flux tubes ( or @xmath0-loops , @xcite ) .
the rise of those loops under the influence of coriolis forces , turbulent convection and large - scale flows ( rotation and meridional circulation ) would then produce the wide variety of large - scale flux emergence observed at the solar surface , with the robust properties summarized above .
several numerical experiments of this last step have been performed using various levels of approximation , from the classical `` thin flux tube '' models @xcite to the more recent two - dimensional and three - dimensional calculations ( see @xcite for a detailed review on this subject ) .
in particular , @xcite showed the necessity to start with a twisted magnetic field to counteract the vorticity generation within the flux tube and its subsequent breakup . @xcite demonstrated the significant effect of convective motions on the deformation of the magnetic loop when the initial field was comparable to the equipartition field strength ( corresponding to a magnetic energy equals to the kinetic energy of the strongest downflows ) .
more recently , @xcite considered the effects of spherical geometry , convective motions , and self - consistently developed mean flows ( differential rotation and meridional flow ) on a rising flux tube .
it was found that an initial twist of the field lines of about 14 turns along the @xmath4 longitude as well as an initial field strength of about @xmath5 were necessary for a coherent radial rise of the tube , that the tilt angle was significantly influenced not only by the initial twist and coriolis force but also by the local convection , and that meridional circulation was able to advect initially weak flux tubes toward the poles when they reached the top of the domain .
they showed moreover a quantitative study of the effect of magnetic diffusivity by varying the prandtl number in their simulations .
@xcite also investigated the effects of spherical geometry but on an @xmath0-loop ( a flux tube made buoyant in a limited range of longitudes ) and thus focused on the properties of simulated individual active regions , but without convection .
we here propose to compute the dynamical evolution of an @xmath0-loop both in an isentropic and fully convective spherical shell , seeking to answer the following questions .
* what are the physical processes at the origin of the observed tilt angles of active regions and is it possible to reproduce joy s law ? * what is the influence of convective motions on a rising @xmath0-loop and the subsequent bipolar emerging regions ?
* can we quantify the effects of large - scale mean flows ? * how important is the twist of the magnetic structures in the evolution within the convection zone and on the morphology of efrs ? *
how does the structure of simulated active regions compare to observations ?
this work is thus an attempt to answer those questions using numerical simulations of a convective spherical shell , using the anelastic spherical harmonic ( ash ) code which solves the mhd equations under the anelastic approximation , valid within the bulk of the convection zone .
we do not address here the evolution of emerging flux in the solar atmosphere which needs to be computed in a fully compressible simulation .
extensive studies of magnetic flux emergence in the atmosphere have been performed ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) which showed that magnetic flux reaching the photosphere could undergo a dynamical expansion into the atmosphere as a result of the non - linear growth of the magnetic buoyancy instability and that it is difficult for a twisted flux tube to rise bodily into a corona because of the heavy plasma trapped in the magnetic dips .
successive magnetic reconnections could however help to achieve actual emergence of twisted magnetic loops into the corona @xcite .
this article is organized as follows .
section [ sect_model ] presents the model chosen , the hydrodynamic background as well as the initial conditions which will favor the emergence of individual active regions .
the following short section [ sect_cases ] presents the various cases calculated , which will be considered in an isentropic background ( without convection ) and with convective motions acting on the magnetic loop .
sections [ sect_isen ] and [ sect_conv ] respectively presents the results of the mhd simulations in the isentropic and convective cases .
finally , we discuss the results and conclude in section [ sect_conclu ] .
all calculations presented in this work were computed with the ash code , which solves anelastic magnetohydrodynamic equations in a rotating spherical shell.ash uses a pseudo - spectral method in space and semi - implicit approach in time ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the full nonlinear evolution of the velocity and magnetic fields is computed , whereas the equations are linearized in thermodynamic variables with respect to a spherically symmetric mean state to have density @xmath6 , pressure @xmath7 , temperature @xmath8 , and specific entropy @xmath9 .
perturbations are denoted as @xmath10 , @xmath11 , @xmath12 , and @xmath13 .
the equations being solved are @xmath14 @xmath15 @xmath16=-{\bf \nabla } p + \rho { \bf g}\\ \nonumber & + & \frac{1}{4\pi}{\bf ( \nabla \times b ) \times b } -{\bf \nabla \cdot \cal d}-[{\bf \nabla}\bar{p}-\bar{\rho}{\bf g } ] , \label{eqns}\end{aligned}\ ] ] @xmath17+\frac{4\pi\eta}{c^2}{\bf j}^2 \\
\nonumber & + & 2\bar{\rho}\nu\left[e_{ij}e_{ij}-\frac{1}{3}({\bf \nabla \cdot v)}^2\right],\end{aligned}\ ] ] @xmath18 where @xmath19 is the local velocity in spherical coordinates in the frame rotating at a constant angular velocity @xmath20 , @xmath21 is the gravitational acceleration , @xmath22 is the magnetic field , @xmath23 is the current density , @xmath24 is the specific heat at constant pressure , @xmath25 is the radiative diffusivity ( fitted to a one - dimensional seismically calibrated solar model , see @xcite ) , @xmath26 is the effective magnetic diffusivity , and @xmath27 is the viscous stress tensor . to represent the unresolved subgrid - scale processes , the code uses an effective eddy viscosity @xmath28 and an effective thermal eddy diffusivity .
they are both here to model the enhanced diffusion associated with turbulent convective motions in the bulk of the convection zone . in the present simulations , they are chosen to vary in radius only as @xmath29 . the thermal diffusion @xmath30 acting on the mean entropy gradient ( contrary to the thermal eddy diffusivity @xmath31 which acts on the entropy perturbation ) occupies a narrow region in the upper convection zone .
its purpose is to transport heat through the outer surface where radial convective motions vanish @xcite and is thus chosen such that the entire luminosity will be transported at the top of the domain by the so - called unresolved heat flux @xmath32 .
the diffusivity @xmath31 is purely dissipative and acts to smooth out entropy variations , whereas @xmath30 is essentially a cooling term near the top of our computational domain . to complete the set of equations
, we use the linearized equation of state , @xmath33 where @xmath34 is the adiabatic exponent ( equal to @xmath35 here ) , and assume the ideal gas law @xmath36 where @xmath37 is the ideal gas constant , taking into account the mean molecular weight @xmath38 corresponding to a mixture composed roughly of 3/4 of hydrogen and 1/4 of helium per mass . the reference or mean state ( indicated by overbars ) is derived from a one - dimensional solar structure model @xcite .
it begins in hydrostatic balance so that the bracketed term on the right - hand side of equation[eqns ] initially vanishes .
however , as the simulation evolves , turbulent and magnetic pressures drive the reference state slightly away from strict hydrostatic balance .
the ash code uses a spectral decomposition for all the variables , namely , spherical harmonics in the horizontal direction and chebyshev polynomials in radius .
the collocation points are thus the zeros of the legendre polynomials in latitude , uniformly distributed in longitude , and the distribution is denser at both ends of the domain in the radial direction . the code integrates the mhd equations presented above over @xmath39 ( with an overall density contrast of about @xmath40 between top and bottom ) , @xmath41 and @xmath42 .
the resolution is chosen here such that at least @xmath43 points in radius and in latitude are contained inside the flux tube section .
the minimal resolution used here is thus @xmath44 .
finally , the boundary conditions for the velocity are impenetrable and stress - free at the top and bottom of the shell .
we impose a constant entropy gradient top and bottom for the isentropic case and for the fully convective case , a latitudinal entropy gradient is imposed at the bottom , as in @xcite . in all cases , we match the magnetic field to an external potential magnetic field at the top and the bottom of the shell @xcite . in this paper , we implement initial conditions for the magnetic field similar to what was used in @xcite .
an initially axisymmetric magnetic structure is embedded in an unmagnetized stratified medium . in order to keep a divergenceless magnetic field
, we use a toroidal - poloidal decomposition , @xmath45 the expressions used for the potentials @xmath46 and @xmath47 for the flux tubes are : @xmath48 \times \left[1+\tanh\left(2\frac{\theta-\theta_t}{a / r_t}\right)\right]\ ] ] @xmath49 \times \left[1+\tanh\left(2\frac{\theta-\theta_t}{a / r_t}\right)\right]\ ] ] where @xmath50 is a measure of the initial field strength , @xmath51 is the tube radius , @xmath52 is the position of the tube center and @xmath53 is the twist parameter . in @xcite , we derived an expression for the winding degree of the field lines ( i.e. , the number of turns that the field lines make over the whole tube length @xmath54 , assuming a uniform twist ) :
@xmath55 where the tangent of the pitch angle @xmath56 is related to the twist parameter via the following expression , considering that we are at @xmath57 and at the tube periphery @xmath58 , as in @xcite : @xmath59 in all cases , the tube radius is set to @xmath60 , about a 10th of the depth of the modeled convection zone and is introduced at the base of the cz at @xmath61 .
if we consider that the typical flux in an observed active region is @xmath62 and that an initial flux tube has a magnetic field of about @xmath63 as in our simulations , the conservation of flux tells us that the radius of the loop should be in reality closer to @xmath64 .
our flux tubes are thus bigger than what could be expected .
different values for the loop radius have been shown to influence flux emergence ( * ? ? ?
* ; * ? ? ?
* see also the discussion of section [ sect_conclu ] ) .
we here limit our parameter study to a fixed loop radius ( see * ? ? ?
* for a different set of parameters ) .
the initial field strength @xmath50 , the initial twist of the field lines @xmath53 as well as the latitude of introduction @xmath65 will be varied in our models to investigate the influence of these various parameters . in order to get a flux tube buoyant on a small portion in longitude only , we initially apply a perturbation on the background entropy field located at the initial position of he flux tube in @xmath66 and possessing a gaussian profile in longitude @xmath67 .
the entropy perturbation has thus the following expression : @xmath68 \times \\ \nonumber & & \frac{2r_t}{a } \times \frac{1}{\cosh^2 \left(2\frac{\theta-\theta_t}{a / r_t}\right ) } \times\\ \nonumber & & \left[-c_s+\exp \left ( -(\frac{\phi-\phi_0}{\phi_{e}})^2\right ) \right]\end{aligned}\ ] ] where @xmath69 is the amplitude of the entropy perturbation , @xmath70 is a constant value controlling the buoyancy of the rest of the loop , @xmath71 is the longitude of maximum perturbation ( or buoyancy ) , and @xmath72 is the extent of the entropy perturbation .
the effect of such a perturbation on entropy is to produce an additional density deficit inside the flux tube at a particular location in longitude . as a consequence
, we can derive the new maximum density contrast between the tube and its surroundings , which will result from this additional perturbation .
total pressure equilibrium is not enforced at the beginning of the simulation and since an initial magnetic tension is also present in most of our cases ( the twisted cases ) , the loop is not initially in mechanical equilibrium outside the portion which is made buoyant , even if @xmath70 is modified .
however , there is an initial adjustment visible at the very early stages of the dynamical evolution where the pressure perturbation inside the loop will tend to compensate for the magnetic pressure , in particular the loop gets slightly squeezed in an alfvn time at the very beginning of the simulation , without much influence on the following evolution .
we thus derive here only an indication for the efficiency of the magnetic buoyancy in a situation of pressure equilibrium and entropy perturbation ( not exactly satisfied here ) in writing the following relations respectively for the total pressure and the entropy : @xmath73 @xmath74 where @xmath75 and @xmath76 are the gas pressure respectively outside and inside the tube and where the maximum of @xmath77 is given by equation [ eq_spert ] and taken at @xmath78 , @xmath57 , and @xmath79 .
we then use an additional relation to make the density @xmath10 appear : @xmath80 with @xmath24 the specific heat at constant pressure ( equal here to @xmath81 ) , @xmath82 the specific heat at constant volume , and @xmath83 the adiabatic index .
these equalities lead to the following relation between entropy , pressure , and density inside and outside the flux tube : @xmath84 and thus we find an expression for the maximum buoyancy of our magnetic structures : @xmath85 where @xmath86 will be the maximum field strength on the tube axis .
a first - order taylor expansion has been done to give an approximate value of the buoyancy , considering that both @xmath87 and @xmath88 are equally small values . in the remaining of the paper
, we will use the following parameters for the standard cases for the entropy perturbation : @xmath89 , @xmath90 , @xmath91 and the average external pressure used to estimate the maximum buoyancy will be @xmath92 .
we use a fixed value for @xmath70 since we do not seek to maintain a perfect mechanical equilibrium for the rest of the loop , otherwise the value of @xmath70 should depend on the magnetic field .
the same hydrodynamical background was used in this work as in @xcite .
namely , two situations will be investigated . in the first
, the magnetic structure is introduced in an isentropic layer with solid body rotation and without convection . in the second situation , the convective instability is triggered and differential rotation as well as a large - scale meridional flow are self - consistently created within the bulk of the convection zone .
the competition between buoyant loops and convective motions and large - scale flows will in this case be studied and compared to the stably stratified situation .
in particular , we focus on the shape of the developing loop , its asymmetries and the morphology of the emerging radial field it generates .
moreover , we focus in this study on the comparison to observations of large - scale efrs , extrapolating the magnetic configurations we get from our simulations at the upper boundary ( i.e. , @xmath93 ) to what could happen higher up , closer to the photosphere .
we specifically investigate the evolution of flux , tilt angle and spot morphology during emergence .
we investigate in this study seven different cases .
all parameters used for each case are summarized in table [ tab ] .
four of the cases go by pair with the same buoyancy ( but different field strength and entropy perturbation : cases 1 & 3 and 2 & 4 ) and the last three are extreme cases ( the weakest cases 0a and 0b and strongest field case 5 ) . for cases 1 , 2 , 3 , and 4 ,
the twist parameter was also varied .
the cases with a right - handed ( or positive ) twist ( corresponding to @xmath94 ) will be called by the case number with the suffix twp , and for the left - handed ( or negative ) twist , we will add twn .
for example , the case where @xmath95 , @xmath96 ( corresponding to a buoyancy @xmath97 of @xmath98 according to equation [ eq_rho ] ) and which has a right - handed twist @xmath94 will be called case 1twp .
some additional cases were also computed : case 1 with the loop introduced at the latitude of @xmath99 : case 1twp60 and @xmath100 : case 1twp15 .
case 3 without twist : case 3notw and with the twist at its threshold value ( defined in the last paragraph of section [ twist ] ) : case 3twth .
cccccccccc case0a & @xmath101 & @xmath102 & @xmath103 & @xmath103 & @xmath104 & @xmath105 & @xmath106 + case0b & @xmath107 & @xmath108 & @xmath103 & @xmath103 & @xmath104 & @xmath105 & @xmath109 + case 1twp & @xmath110 & @xmath111 & @xmath103 & @xmath103 & @xmath112 & @xmath113 & @xmath98 + case 1twp15 & @xmath110 & @xmath111 & @xmath103 & @xmath114 & @xmath112 & @xmath113 & @xmath115 + case 1twp60 & @xmath110 & @xmath111 & @xmath103 & @xmath116 & @xmath112 & @xmath113 & @xmath115 + case 1twn & @xmath110 & @xmath111 & @xmath117 & @xmath103 & @xmath112 & @xmath113 & @xmath115 + case 2twp & @xmath118 & @xmath113 & @xmath103 & @xmath103 & @xmath112 & @xmath113 & @xmath119 + case 2twp15 & @xmath118 & @xmath113 & @xmath103 & @xmath114 & @xmath112 & @xmath113 & @xmath119 + case 2twp60 & @xmath118 & @xmath113 & @xmath103 & @xmath116 & @xmath112 & @xmath113 & @xmath119 + case 2twn & @xmath118 & @xmath113 & @xmath117 & @xmath103 & @xmath112 & @xmath113 & @xmath119 + case 3twp & @xmath118 & @xmath113 & @xmath103 & @xmath103 & @xmath104 & @xmath105 & @xmath115 + case 3notw & @xmath118 & @xmath113 & @xmath120 & @xmath103 & @xmath104 & @xmath105 & @xmath98 + case 3twth & @xmath118 & @xmath113 & @xmath121 & @xmath103 & @xmath104 & @xmath105 & @xmath122 + case 3twn & @xmath118 & @xmath113 & @xmath117 & @xmath103 & @xmath104 & @xmath105 & @xmath98 + case 4twp & @xmath123 & @xmath124 & @xmath103 & @xmath103 & @xmath104 & @xmath105 & @xmath125 + case 5twp & @xmath126 & @xmath127 & @xmath103 & @xmath103 & @xmath112 & @xmath113 & @xmath128 [ tab ] the values of the entropy perturbation were chosen to ensure a rise time of the same order of the convective turnover time , i.e. , between 10 and 40 days approximately .
we can compare the buoyancy of the loops with and without the entropy perturbation by setting @xmath129 in equation [ eq_rho ] .
we find that the value of @xmath130 is multiplied by @xmath131 when the entropy perturbation is introduced in case 1 ( compared to the case without entropy perturbation ) and by a factor @xmath132 in case 5 , so that the stronger field cases are less influenced by the entropy perturbation applied .
moreover , the values of the entropy perturbations used here are comparable in amplitude to the typical fluctuations measured at the base of our computational domain in our convective runs .
we first compute the evolution of a single @xmath0-loop evolving in a stable background with respect to convection . it will first confirm that we indeed correctly control the buoyancy of the flux tube with the initial conditions we apply .
secondly , it will help us see if the important parameters for the evolution of a uniformly buoyant flux tube are still the same here and what differences may be pointed out . we know from previous experiments that a certain amount of twist is needed for a magnetic flux tube to be able to rise coherently in a convectively stable ( or unstable ) background ( e.g. , * ? ? ?
indeed , in a non - twisted case , the azimuthal vorticity created by the gravitational torque acting on the loop can not be compensated .
two counter vortices thus quickly appear and break the loop apart , preventing it from emerging at the surface .
it was argued previously that this situation may differ when an @xmath0-loop is considered , instead of a uniformly buoyant structure @xcite .
the additional magnetic tension coming from the deformation of the magnetic torus may oppose the vorticity generation .
-loop of case 3notw , introduced at a latitude of @xmath133 , at two instants in the evolution , @xmath134 and @xmath135 days .
magenta contours of the azimuthal component of the magnetic field are superimposed to the azimuthal vorticity which presents a characteristic profile of two counter vortices.,width=283 ] figure [ figure_notw ] shows the section of an untwisted @xmath0-loop corresponding to case 3notw of table [ tab ] .
the zoom is made on the longitude where the loop is caused to be buoyant , thanks to the combined effects of magnetic pressure and entropy perturbation .
two instants in the evolution are shown .
we represent here the contours of the azimuthal ( or axial ) component of the magnetic field superimposed on the azimuthal vorticity .
it is rather clear in this case ( especially on the left panel ) that strong vortices are formed inside the flux tube , which have the tendency to significantly deform the magnetic structure .
however , as the loop rises further up , it becomes more arched and more magnetic tension is generated . as a consequence , even without any twist ,
some coherence is maintained in this case , even if a large amount of flux is lost in the process . ) .
the three cases represented are the untwisted case ( @xmath136 ) case 3notw , the twisted case case 3twp ( @xmath94 ) , and the intermediate case case 3twth where the twist value is at its threshold ( @xmath137).,width=283 ] this flux loss is visible in figure [ figure_notw2 ] where the profile of the toroidal magnetic field is shown at the radius where it peaks after about @xmath138 days of evolution , as a function of colatitude .
we remind that the flux ropes were introduced here at @xmath99 of colatitude and with an amplitude of @xmath139 . in the non - twisted case ( dashed line on the figure )
, we clearly see that the two counter vortices have created two distinct magnetic flux concentrations , whose maximum amplitude reaches a value of @xmath140 while the maximum amplitude of the twisted case is @xmath141 .
the flux rope has been eroded by the vorticity generation and the two vortices tend to dominate the further evolution of the magnetic structure which separates into two flux concentrations . however , along the loop ( following its axis ) , this double - peak structure is not as clearly visible as at the longitude we chose here .
this feature may be the consequence of the asymmetry which develops between the leading and trailing leg .
we will come back on this in the next section .
we decided to define the twist threshold as the value above which the double - peak structure is not present anymore at any longitude . with this criterion
, we find that the twist threshold for this case is @xmath137 , corresponding to @xmath142 turns along the @xmath100 of extension of the @xmath0-loop .
we note that this threshold value is reduced by a factor @xmath143 compared to the uniformly buoyant flux tubes of @xcite .
this is due to the additional tension coming from the arching of the @xmath0-loop in this case , which has already the effect of limiting the vorticity generation inside the magnetic structure .
this result is in agreement with the work of @xcite .
the case where the twist is set at its threshold value is also represented in figure [ figure_notw2 ] ( dotted line ) . in this case , the maximum value of the field is reduced compared to the more twisted case but the coherence of the loop is kept . the buoyancy force is proportional to @xmath144 which is given by expression ( [ eq_rho ] ) . while the loop rises because of the buoyancy force , a retrograde longitudinal flow develops in order to conserve angular momentum ( see section below ) . as a consequence , a coriolis force oriented toward the rotation axis and perpendicular to this axis also develops , opposing the component of the buoyancy force perpendicular to the rotation axis .
the only uncompensated force is then parallel to the rotation axis and oriented toward the pole .
if the buoyancy force is not strong enough , the loop will thus have the tendency to rise parallel to the rotation axis and emerge at a much higher latitude than where it was introduced .
this is the case at fixed @xmath145 when the magnetic energy is not strong enough .
we recover this situation when the poloidal field is too weak ( corresponding to a low twist value , enough to maintain coherence of the loop but not enough to make it rise radially ) . at @xmath146 days at the longitude of maximum buoyancy for case 0a .
the trajectory of the loop is here clearly parallel to the rotation axis.,width=283 ] figure [ figure_vertical ] is an illustration of a case where the rise is mainly parallel to the rotation axis . in this case
0a , the twist was set to @xmath94 but the initial axial field strength was set to @xmath147 , and the entropy perturbation to the rather weak value of @xmath148 , corresponding to a density perturbation of @xmath149 . in this case , after @xmath114 days of evolution , the trajectory of the loop is clearly parallel to the rotation axis , since the coriolis force has strongly acted on the magnetic field .
indeed , as shown by @xcite in the framework of the thin flux tube approximation , while the loop rises , it develops a retrograde longitudinal flow in its interior .
a coriolis force is associated with this longitudinal flow which tends to oppose the outward component of the buoyancy force perpendicular to the rotation axis , leaving only the component parallel to the rotation axis to act on the ring .
interestingly , @xcite find that in the adiabatic cases ( similar to what we are studying in this section ) , the transition between vertical and radial rise happens between @xmath150 and @xmath151 .
we also find here that cases 0a and 0b , where @xmath97 is around @xmath152 , tend to be strongly influenced by the coriolis force while the stronger cases , where @xmath153 , exhibit a much more radial trajectory .
we note that due to the low value of the buoyancy used for case 0a , the rise of the loop is rather slow ( about 40 days are needed to reach the top of our domain ) . in order to ensure an evolution dominated by the dynamical effects and not only by diffusion
, we reduced the magnetic diffusivity from @xmath154 to @xmath155 , and thus used a magnetic prandtl of 5 instead of 1 as with all the other cases .
it does not change the fact that the rise is parallel to the rotation axis in this case , it is just to compute a case where the rise time is significantly lower than the diffusive time ( which would be of order @xmath156 days for the @xmath157 case and thus too close to the rise time of @xmath158 days ) .
the same kind of evolution parallel to the axis of rotation is found when stronger tubes are introduced with a lower twist . indeed , if the twist is low as we saw before , the amplitude of the magnetic field inside the loop quickly decreases until it reaches a value where the buoyancy force is not strong enough to keep the radial trajectory .
this is visible on the right panel of figure [ figure_notw ] where the initial axial field was @xmath139 but the initial poloidal field was @xmath120 . to correctly define a minimum value for a radial rise here ,
it is thus necessary to take into account the three components of the magnetic field rather than focus on the axial field . let us consider a case where the entropy perturbation is the weakest , corresponding to a slow rise and then to a maximum effect of the coriolis force .
we then find that for a twist parameter @xmath94 ( about 1 turn over the extension chosen here ) , the initial magnetic field needed for a radial rise should be at least of @xmath63 .
it can be argued that this value is strong compared to what could be expected at the base of the convection zone from dynamo action but it could be reduced if a stronger entropy perturbation was considered , which is locally possible in the turbulent environment existing at the base of the convection zone .
+ to summarize the two previous subsections , we can say that to maintain a coherent radial rise of our @xmath0-loop in our simulation , an initial twist of @xmath142 turns along the extension of the loop , as well as an initial field strength of @xmath159 are necessary .
we note that those two values are reduced compared to the uniformly buoyant calculations .
this is promising since the observed twist values are of the order of less than 1 turn across the active region @xcite and typical fields of more than @xmath159 are difficult to produce in present dynamo calculations @xcite .
we now study the evolution of a sufficiently twisted and strong loop so that it will maintain its coherence while it rises radially .
the loop is introduced at @xmath133 in latitude , with a maximum buoyancy ( i.e. , entropy perturbation ) located at @xmath160 in longitude , with a twist parameter @xmath94 representing about @xmath161 turns around the tube axis over the whole @xmath4 and thus corresponding to about @xmath162 turns over the extension of the buoyant part ( @xmath90 ) . with this twist parameter ( as we saw before ) , the loop rises radially for a field strength of @xmath63 .
figure [ figure_isen3d ] shows a volume rendering of the toroidal field evolution in case 1twp .
the asymmetry between the trailing and leading legs develops during the evolution and is clearly seen in the last two panels of the figure , at @xmath163 days and @xmath164 days .
this section will be devoted to the study of this asymmetry and the influence of buoyancy and field strength .
the action of rotation and the associated coriolis force over the whole loop will also be detailed here .
this process will lead to the appearance of a tilt of the axis with respect to the east - west direction and will also depend on the magnetic field strength , as we will see in the following section .
let us enter now into the details of the asymmetry between the trailing and leading legs and the effect of the field strength on it .
this asymmetry , already seen in thin flux tubes calculations ( see * ? ? ?
* ) , is due to the difference in azimuthal velocity between the top of the loop and its feet . indeed , since the flux tube needs to conserve its angular momentum @xmath165 , a retrograde flow with respect to the rotating frame will develop inside the loop while it rises ( i.e. , while @xmath166 increases ) .
since the apex of the loop is higher than the feet which stay attached to the bottom of the convection zone , a stronger retrograde flow appears at the apex .
the profile of the flow along the magnetic structure then leads to the inclination asymmetry of the trailing and leading legs . in some of our cases , since the parameter @xmath70 is always kept the same , parts of the loop located outside the buoyant portion can be slightly denser than their surroundings . as a consequence
, those regions may sink toward the base of the convection zone and favor the difference in azimuthal velocity causing the asymmetry .
however , in most cases ( except cases 0a , 0b , 1twp , and 1twn ) , the magnetic buoyancy can overcome this effect and the rest of the loop is slightly positively buoyant , thus on the contrary reducing the asymmetry . however , this explanation is valid as long as the lorentz forces do not dominate the flow evolution inside the loop .
figure [ figure_eqsl ] shows a snapshot of the magnetic energy and the azimuthal velocity for four different cases : cases 0b , 1twp , 2twp , and 5twp .
we note here that case 0b has a magnetic field strength of @xmath141 and is thus still significantly deflected toward the pole but is an interesting case because it maximizes the asymmetry between the legs of the loop .
the snapshots were chosen so that all loops have reached the same height ( namely @xmath167 ) , which corresponds to different physical times since the buoyancy force ( i.e. , the @xmath144 ) is different in all cases . in order to quantify the asymmetry , we measure the angle between the tangent at the apex and the line going through the middle of the two footpoints and the apex .
those lines are shown in magenta in the top panels of figure [ figure_eqsl ] .
we find that the angles have the following values : @xmath168 , @xmath169 , @xmath170 , and almost @xmath171 ( meaning no asymmetry in this case 5twp where the magnetic field and the buoyancy force are the strongest ) .
we thus note that the asymmetry strongly decreases from case 1twp to case 2twp , even if the density deficits are initially not very different .
the reason for this jump in the asymmetry is the difference in field strength ( @xmath139 for case 2twp against @xmath172 for case 1twp ) .
indeed , in this case , the lorentz force starts to become important in the generation of azimuthal velocity : strong values of @xmath173 appear at the location of strong gradients of magnetic field .
this is much more obvious in the stronger case 5twp .
the profile of the azimuthal velocity shown on the last snapshot of the lower panel is clearly related to the gradients of magnetic field and the retrograde flow due to the conservation of angular momentum is so weak compared to the magnetic effects that it is not visible anymore .
instead , an azimuthal velocity directed toward the bisector of the tube appears on both sides of the loop ( indicated by black arrows in the last panel ) , resulting in a symmetric arched structure ( visible in the top right snapshot ) with both feet showing an almost vertical orientation .
if we write down the expression of the lorentz force involved in the evolution of @xmath173 , we have : @xmath174 this equation shows that the regions of strong magnetic field gradients will correspond to the regions of production of @xmath173 , which can be verified from the last panel of figure [ figure_eqsl ] , where the concentrations of @xmath173 are located mainly at the footpoints ( where the gradients of @xmath175 and @xmath176 with respect to @xmath67 are strong ) and higher up at the loop periphery ( where the gradients of @xmath177 with respect to @xmath166 and @xmath178 are strong ) . to be more precise
, we can follow the evolution of the maximum and minimum of the azimuthal velocity which develops during the rise of the flux tube from the base of the convection zone toward the surface . in figure [ figure_eqsl2 ]
, we plot the maximum and the minimum of @xmath173 , in order to understand the differences in the asymmetries seen in figure [ figure_eqsl ] . for each case , the maximum and minimum azimuthal velocities are measured over the whole domain but those extreme values are always reached inside or in the vicinity of the magnetic structures .
moreover , the various cases ( except for case 0b ) only differ by the initial magnetic field strength of the loop and the differences in @xmath179 will thus be directly related to the differences in the magnetic properties .
the most striking feature is the strong increase of the maximum velocity in all cases .
this sharp increase happens in all cases when the loop reaches the mid - cz , i.e. , when the loop starts to get a significant curvature .
if we look at equation ( [ eq_vphi ] ) , we can understand that this sharp increase is due to the gradients with respect to the @xmath67-coordinate of @xmath175 and @xmath176 which appear only when the loop becomes arched , i.e. , after a certain rise time , contrary to the gradients of @xmath177 with respect to @xmath166 and @xmath178 , which exist at the beginning of the simulation already .
if we now turn to look at the minimum of @xmath173 , we realize that this change of slope is also visible at the same time but only for the strong field cases .
they are indicated by the red arrows in panels ( @xmath180 ) and ( @xmath181 ) . for the retrograde flow with respect to the rotating frame indeed ,
another ingredient plays a role , the conservation of angular momentum . as we said before
, it will have a tendency to produce a negative @xmath173 which is dominant in the weak - field cases ( panels ( @xmath51 ) and ( @xmath182 ) ) but dominated by the lorentz force terms in the strong field cases ( panels ( @xmath180 ) and ( @xmath181 ) ) . as a consequence , strong gradients of the magnetic field and the associated magnetic tension
will have the tendency to oppose the deformation of the loop while it rises and it explains why the asymmetry between the two legs of the loop is reduced when the initial field is strong enough ( more than @xmath139 ) .
this feature was already observed in thin - flux tube calculations ( e.g. , * ? ? ?
* ; * ? ? ?
* ) and a similar argument involving the stiffness of intense fields was used to explain the weaker asymmetry seen in strong field cases .
we note moreover that since the rise time of strong tubes is smaller , the coriolis force will have less time to act on the loop and the asymmetry will also be reduced because of shorter rise times .
going back to figure [ figure_isen3d ] , we see that when the loop forms and rises , some negative toroidal field is created ( blue colors in the figures ) , while the initial toroidal field was purely positive . this can be explained by a local @xmath0-effect :
the azimuthal velocity gradient built at the periphery of the loop while it rises will shear the small poloidal field component present at the tube boundary mainly .
this @xmath0-effect , related to a shearing of poloidal field is one of the main ingredients of the dynamo mechanism thought to be responsible for the generation of magnetic field in differentially rotating astrophysical objects ( see for example @xcite for a review on these processes in the solar context ) .
the presence of such an effect could be surprising , especially in the untwisted case without convection , since no poloidal field nor azimuthal velocity gradients exist at the initial stage .
however , while the loop forms , the magnetic field is no longer purely toroidal since the field lines get curved and will thus get a mainly radial component . at the same time , gradients of azimuthal velocity will be created as shown above .
both these effects are then favorable to get an @xmath0-effect which will be able to create additional toroidal field of both signs .
we now turn to investigate the characteristics of the individual bipolar regions emerging at the top of our computational domain in all situations . in particular , we focus on the tilt angle , the asymmetry in the field strength in the two spots and the morphological properties of the regions which can be related to observations .
we decide here to focus on four cases : cases 1twp and 1twn and cases 3twp and 3twn .
the main difference between the two cases with right - handed twist is the field strength which is multiplied by 2 in case 3 but buoyancy is kept the same through a change of initial entropy perturbation .
moreover , we know from previous studies ( e.g. , * ? ? ?
* ) that the sign of the initial twist will have a dramatic effect on the tilt angle of the emerging region .
we address here this question by computing cases 1 and 3 with a right - handed as well as left - handed twist .
figures [ figure_shsl0 ] and [ figure_shsl ] show a zoom on the emerging region in both cases by the time the apex of the loop reaches the top of our domain ( @xmath183 ) .
blue indicates negative radial field and red positive .
the first thing we note is the typical evolution of the inclination of the bipolar structure .
starting from north / south because of the twist of the loop apex , it becomes more and more east / west as we see the legs of the loop emerging . in agreement with previous thin - flux tube calculations ( e.g. * ? ? ?
* ) , we find that a stronger initial field produces a smaller tilt angle since the magnetic tension will tend to oppose the tilting of the magnetic structure .
this effect is visible when comparing the left panels of figures [ figure_shsl0 ] and [ figure_shsl ] which differ by a factor 2 in initial field strength .
the tilt angle of the bipolar structure and the effect of the initial twist is consistent with previous simulations , especially of @xcite . indeed , in the left - handed case ( which is thought to be dominant in the northern hemisphere ) , the tilt angle is not in agreement with observations which tend to show an inclination of the bipolar spots much closer to what we get in the right - handed twist .
it was argued previously @xcite that to reconcile simulations with observations , a very weak left - handed twist had to be introduced so that the coriolis force acting on the loop would be strong enough to change the tilt angle from positive to negative sign .
the problem is that such a weak twist , not sufficient to maintain a full coherence of the magnetic structure , leads to a strong flux loss and such structures would be likely to have difficulties to rise in a convective environment .
if we look at figure [ figure_shsl ] where the initial field strength is decreased to @xmath63 but the amount of twist kept the same , we manage to get a case where the initial left - handed twist is strong enough to maintain the coherence of the loop and where the tilt angle of the emerging radial field gets in better agreement with observations .
the coriolis force has been able to act sufficiently on this loop to rotate the two opposite polarities clockwise and produce a final tilt angle of the same sign in both cases . to illustrate this ,
figure [ figure_tilt ] shows on the same plot the temporal evolution of the tilt angle measured in cases 1twp and 1twn , at @xmath184 .
the tilt angle was here determined by locating the peak values of each polarity and measuring the angle between the line linking them and the east - west direction . at the beginning of emergence ,
the angle is close to @xmath171 for both cases , but with opposite signs .
as the emergence progresses , the absolute value of the tilt reduces in both cases . at this stage , the effect of the twist is dominant and corresponds to what is measured here .
after about 14 days of evolution , the rapid change of tilt stops and the effect of the coriolis force starts to be visible , especially on case 1twn . as a consequence , the curve in case 1twp seems to saturate around the value of @xmath185 and starts to increase again slightly because of the rotation of the spots , to reach a value of about @xmath99 after 26 days of evolution . in case 1twn
, the tilt angle continues to increase and crosses the @xmath186 line after about 15 days . at this instant ,
a change of sign has occurred and the tilt has thus become compatible with observations . a final value of about @xmath100 is reached after @xmath187 days when diffusion then starts to act , which lies in the acceptable range of values when compared to observations at this latitude .
.,width=283 ] we now turn to study the influence of latitude on the tilt angle . as we expect and as seen in figure [ figure_shsl4 ] , the tilt angle depends on the initial latitude since the coriolis force depends on the latitude . from the estimate of @xcite ,
the tilt angle is a decreasing function of the cosine of the colatitude ( or an increasing function of latitude ) , as seen in figure [ figure_shsl4 ] and as seen in observations . as we saw before in the left - handed case
, the bipolar spot will have the tendency to rotate clockwise because of the coriolis force acting differently on the two legs of the loop . as a consequence ,
the orientation of the bipolar spots in the case where it was introduced at high latitudes ( @xmath99 ) is close to north / south , and not only because we only see the twist of the field lines here but rather because the axis of the loop has been rotated clockwise by about @xmath170 . when the initial latitude is @xmath100 , the typical tilt angle we get is around @xmath188 while it increases to @xmath99 when a loop at initially @xmath133 is considered .
moreover , we note on this figure that the emergence occurs at slightly different locations in longitude .
the center of the bipolar structure being located at around @xmath170 in case 1twp15 and @xmath171 in case 1twp60 .
this is due to the asymmetry of the loop which is much more pronounced in the low latitude cases since the action of the coriolis force on both legs of the loop is stronger then .
( case 1twp15 ) , @xmath133 ( case 1twp ) and @xmath99 ( case 1twp60 ) .
the northern hemisphere is represented centered at @xmath171 in longitude and @xmath168 in latitude.,width=283 ] we note here that the values for the tilt angle may appear large compared to the classical observations ( between @xmath189 and @xmath3 typically ) and to previous thin flux tube calculations @xcite but as shown before ( on figure [ figure_tilt ] ) , an initial left - handed twist , associated with a slow rise allowing the coriolis force to act on the loop , is able to produce tilt angles of about @xmath100 in a case where the initial latitude was @xmath133 .
this , in turn , is very compatible with observations at the photosphere .
it is possible here to make such a comparison if we make the reasonable assumption that the coriolis force will not have the time to modify the loop orientation and thus the tilt angle during the fast rise through the last few percents of the convection zone .
we shall keep in mind those results for the convective cases where the spherical shell will now be differentially rotating .
since the tilt angle is also a function of the local rotation rate , the tilt angles we will get in the convective cases may well be different .
if we concentrate again on figure [ figure_shsl0 ] , we immediately notice the typical `` tongue''-shape of the emerging region , especially at a deeper radius @xmath190 .
they are indicated by red and blue arrows in the bottom left panel of figure [ figure_shsl0 ] .
those elongated structures , also seen in the observations @xcite , in cartesian simulations @xcite and studied in details in @xcite , are mainly due to the twist of the field lines showing in the map of the radial magnetic field .
in particular , the orientation of the tongues ( or the direction in which they extend ) give an indication of the sign of the twist .
in the lower left panel of figure [ figure_shsl0 ] , we thus see that both positive and negative polarities extend on their left side when it is the opposite for the lower right panel where the sign of the twist has changed .
another interesting feature here is the appearance of sharp structures around the emerging region , forming an annular shape that we will from now on call a `` magnetic necklace '' .
this annular shape of emerging active regions is also often seen in observations but has never been reported as identified magnetic structures , different from the so - called tongues .
this is true for observations with experiments and instruments as varied as the flare genesis experiment @xcite , _ hinode _ @xcite or _
soho_/mdi @xcite .
those structures are clearly visible in the lower panels of figure [ figure_shsl0 ] but do not really show in figure [ figure_shsl ] because of the low field strength and the rather large magnetic diffusivity in our simulation .
indeed , since those structures are organized at rather small scales , it is necessary to have low enough diffusivities to be able to see those structures at the time of emergence and during their rise . as a consequence
, we decided to compute case 1twn with a lower magnetic diffusivity ( we choose @xmath191 here ) to see if these small - scale structures could be recovered in this case . indeed , as shown on figure [ figure_neck ] , the magnetic necklace is visible on the zoom in the @xmath192 plane on the emerging region at @xmath167 and is also seen in a meridian cut of the loop during its rise ( see red and blue arrows in the left panels ) .
they consist in sharp elongated structures ( not to be confused with the tongues ) lying at the periphery of each polarity and of the same sign that create a ring inside which the two main regions of strong radial field concentrate .
if we compare the regions of appearance of these sharp structures around the bipolar spot with the norm of the vorticity ( or enstrophy ) , it is rather clear that those structures correspond to a velocity shear which produces a strong vorticity . as shown and studied in @xcite , a sharp interface is indeed formed at the boundaries of the flux rope while it rises .
this region is the site of vorticity generation via the presence of a magnetic field during the whole loop evolution .
another source of vorticity exists within the tube interior and is linked to the gravitational torque applied between the tube center and its periphery .
this process creates the counter - rotating vortices which can completely break up the tube during its rise if the twist ( and thus the magnetic tension ) is not strong enough , as seen in section [ twist ] .
looking at the four panels of figure [ figure_neck ] , we understand that the necklace , associated with vorticity generation at the loop edges , will be visible only after a significant part of the loop has emerged . indeed , since the main vorticity generation is located at the sides of the loop , we need to wait until the axis of the loop crosses the surface @xmath193 before being able to observe the sharp magnetic structures . by that time , the tongues have thus already significantly faded away and reconcentrated around the two polarities .
we can thus expect to see those structures in actual observations only after a certain amount of time after the first signs of emergence .
we end this section by noting another difference between cases 1 and 3 . in case 3
, the loop seems to maintain more coherence while it rises , both polarities are well - separated , even at the end of the simulation when the bipolar region reaches its maximum extension .
this is less the case for the weaker field of case 1 .
indeed , both polarities seem to get mixed and the radial magnetic field lines get more easily advected by the velocity field created by the loop itself ( we remind that there is no convection in this case ) . at the end of the simulation
, the magnetic field gets advected at the boundary of the positive radial velocity pattern it creates and finally diffuses away .
if we now focus on this more realistic case 1 ( with a field strength of @xmath172 , and a typical rise time of more than 10 days ) , we can move to a more quantitative study of the evolution of the intensity of the radial field emerging in each spot and of the amount of flux it represents .
the results of this study are shown in figure [ figure_shsl3 ] .
panels ( @xmath51 ) and ( @xmath182 ) show the evolution , during emergence at @xmath193 of the radial field and radial flux in case 1twp and panels ( @xmath180 ) and ( @xmath181 ) in case 1twn . in both cases at the beginning of emergence , the negative field in the leading polarity dominates over the trailing polarity ( by about @xmath194 ) , while the amount of flux is the same in both polarities ( since all emerging field lines of one polarity connect to the other one in the domain of integration ) .
this suggests that the leading polarity is stronger in intensity and more concentrated and the trailing polarity is weaker and less coherent , which is in agreement with observations of emerging active regions ( e.g. , * ? ? ?
moreover , this feature was also first seen in cartesian simulations and explained in @xcite by the evacuation of plasma out of the leading side of the loop into the following side .
the leading side then shrinks and the magnetic pressure increases to balance the same external pressure , resulting in a more intense field in the less extended leading polarity . as @xcite points and as discussed also in @xcite , the anchoring of the field lines outside the buoyant part of the loop is a key ingredient to produce a strong asymmetry in field strength .
it is also what we find here in case 1twn , for example , where a strong asymmetry in inclination develops because of the anchoring of the field lines outside the most buoyant part and where we find a leading leg about 1.4 times more intense than the trailing leg on average ( see panel ( @xmath180 ) of figure [ figure_shsl3 ] .
we note that the values of the flux are quite strong compared to observations ( which report values of about @xmath195-@xmath196 mx ) .
this is due to the scales of our magnetic structures which overestimate the size of a flux tube initially located at the base of the convection zone .
the initial flux in the magnetic structure in case 1 is indeed of @xmath197 .
simulations with flux tubes of radii about @xmath121 times smaller , numerically very challenging , would lead to values for the fluxes much closer to observations @xcite .
it is however important to note that despite the discrepancies between the absolute values of the fluxes , due to numerical limitations , the shape of the temporal evolution of the fluxes is in good agreement with observations .
indeed , the flux in each polarity of the bipolar active regions is observed on mdi magnetograms for example @xcite to increase sharply in the first days of emergence , before saturating and slowly starting its diffusive decay .
we get exactly the same kind of evolution in our simulations , visible especially in panel ( @xmath182 ) of figure [ figure_shsl3 ] .
the flux strongly increases from @xmath120 to @xmath198 in the first 5 days before saturating around this value . in the figure and especially in panel ( @xmath181 ) , we distinguish first a slow rising phase between 8 and 9 days followed by the very sharp increase of the amplitude of the flux after @xmath121 days of evolution .
the first phase is related to the very beginning of the emergence when only the apex of the loop is visible and the only contributor to the flux is the radial magnetic field coming from the twist of the field lines . as the emergence proceeds , however , the feet of the loops where the strong radial magnetic field lies start to emerge and form the bipolar region .
it is thus this fast emergence phase of the arched structure which produces the sharp increase of the simulated ( and possibly of the observed ) magnetic fluxes . as far as the magnetic field strength is concerned
, we can wonder what typical horizontal field strength we would get at the photosphere , starting with our flux tubes located at the base of the cz , in the same way as @xcite . if we follow the evolution of the axial field intensity in time ( and thus in radius ) , we find that the scaling with the reference density is close to @xmath199 ( similar to what is found in @xcite ) , corresponding to an expansion of the loop in the directions transverse to its axis . at the top of our domain , the typical toroidal ( or horizontal ) field is of the order @xmath200 ( for case 1twp for example ) and the density @xmath201 of the order @xmath202 . as a consequence ,
if we consider a density of @xmath203 at the photosphere , the horizontal field strength would reduce to @xmath204 , much weaker than the typical observed values ( around @xmath205 ) .
however , the expansion in the last few percent of the convection zone might be modified ( becoming mostly horizontal , for example , due to a decrease of the plasma @xmath206 ) and then the scaling with density could change drastically and become @xmath207 as shown in @xcite .
we should then keep in mind that we are not modeling the uppermost layers of the convection zone and that it is not relevant here to relate directly the field strengths in our bipolar regions to the field strengths which would be reached at the photosphere . comparing the left - handed and right - handed twist cases
, we find that the evolutions , especially for the radial field , are slightly different . indeed , the dominant leading polarity at the beginning of emergence in case 1twp ( panel ( @xmath51 ) ) saturates at the value of @xmath208 and then drops to the value of @xmath209 while the trailing polarity continues to increase in intensity . at @xmath210 days , the positive trailing polarity thus becomes dominant , before decreasing at @xmath211 days .
this is not the case for case 1twn ( panel ( @xmath180 ) ) , where the leading polarity always stay dominant in strength and approximately equal to the trailing polarity in flux .
this could be due to the asymmetry built up during rise , which is smaller in the left - handed twist case .
this is indeed seen on cuts at constant latitude as in figure [ figure_eqsl ] ( not shown here ) . as a consequence , in case 1twp , the asymmetry between the legs is strong , the leading polarity reaches its maximum strength and thus starts to diffuse away before the trailing polarity .
this explains the decrease in amplitude in figure [ figure_shsl3 ] first seen for the negative field and then for the positive field . in case 1twn , the peak positive and negative radial fields
are measured at @xmath184 at approximately the same time since the asymmetry is less pronounced and they both saturate at the same time ( at the value of @xmath212 for the positive polarity and @xmath213 for the negative one ) , before starting their diffusion .
we now turn to investigate the behavior of a loop similar to the cases presented above , but inside a convective background . as in the global flux tubes simulations
@xcite , the buoyancy force will compete with the downflows to drive the loop towards the top of the computational domain and the emerging region will have significantly different characteristics than in the isentropic cases .
figure [ figure_conv3d ] shows the typical three - dimensional evolution of an @xmath0-loop evolving in a convective environment , this corresponds to case 2twp .
we only see here the evolution of the toroidal field but a turbulent convective velocity field exists all around it .
moreover , differential rotation and meridional circulation have well - established profiles here , as stated in section [ sect_hydro ] .
we recover some aspects of the cases studied in section [ sect_isen ] , like the asymmetry between the legs and the creation of a negative toroidal field at various regions at the loop periphery . again , a local @xmath0-effect is at work here , with a slightly more complex structure since the velocity field is now organized at different spatial and temporal scales .
but the main ingredients ( azimuthal velocity gradients and radial and latitudinal components of the magnetic field ) are still the same and able to locally create negative toroidal field . in such fully convective
setting , we expect the initial field strength in the flux tube to be an important parameter for the evolution and rise of the magnetic field .
indeed , as shown in previous work involving a uniformly buoyant flux tube @xcite , the downflows and upflows of the convective layer control the rise velocity of the tube and are sometimes even able to pin it within the convection zone , resulting in structures emerging at specific longitudes . here again , we get the same kind of competition between convection and magnetic buoyancy , which can be seen on figure [ figure_eqslc ] .
this figure shows the evolution of the loops of case 1twp and case 2twp .
they correspond to the second and third panels of figure [ figure_eqsl ] , for which the asymmetry was clearly modified when the parameters were changed such as to have more intense field . in case 1 , the initial field strength is @xmath172 , i.e. , slightly below the equipartition field strength of @xmath214 , corresponding to the magnetic field in equipartition with the strongest downdrafts at the base of the cz . in case 2 ,
the initial magnetic field strength is @xmath215 , corresponding to @xmath216 times the equipartition field .
we superimpose here the magnetic energy contours to the background radial velocity ( representing the convective motions ) . for figure [ figure_eqsl ]
, we chose to represent the loops at about @xmath217 days since their position in the bulk of the cz was not very different for the isentropic cases . here , at about @xmath217 days , the loops have not reached the same height , the apex of the loop of case 2twp has reached @xmath184 after about @xmath218 days already , whereas the loop of case 1twp reaches the same height at @xmath219 days and was thus significantly slowed down compared to the isentropic case .
this can be explained by the structure of the flow in which the loops are embedded . in case 2twp and as seen especially on the panel at time @xmath220 days , the velocity field associated to the lorentz force is dominant . at the location of maximum magnetic field ,
a strong upflow thus develops ( a maximum positive radial velocity of @xmath221 is reached here ) , the loop is dragged upward and the influence of the downflow shown by the black arrow is limited .
the rise of the loop is thus widely promoted by the strong upward convective cell produced by the intense magnetic field . on the contrary , in case 1twp where the magnetic energy content ( and thus the buoyancy ) is weaker , the radial velocity created by the presence of a magnetic field is of the same order as the background convective motions and the downflows ( as the one indicated by the arrow ) are much more efficient at maintaining the magnetic loop under the surface and preventing the axis from emerging .
another difference we note here compared to the isentropic cases is the degree of asymmetry between the trailing and leading legs .
the strong asymmetry visible in case 1twp when the loop was embedded in an isentropic background is still present at @xmath222 days but less obvious since the advection by the convective motions has also mixed the field lines and significantly modified the structure .
however , in case 2twp at @xmath223 days , the same degree of asymmetry as in the isentropic case is found , showing again that this regime is dominated by the lorentz force , contrary to case 1twp where the magnetic structure is essentially frozen in ( advected by ) the convective motions . in order to gain a better insight on why we get such a different behavior between what we will now call the _ convection dominated _ regime of case 1twp and the _ magnetic field dominated _ regime of the other cases , we plot the evolution of the density deficit inside the loop at its apex as a function of time in cases 1 and 3 .
the results are shown in figure [ figure_density ] .
they correspond to cases 1twn and 3twn but the results are similar for the right - handed twist cases .
our results here and especially the distinction between convection and magnetic field dominated cases are similar to what is obtained in the recent thin - flux tube simulations of @xcite .
indeed , they find that an initial field strength above @xmath139 ensures an evolution dominated by magnetic buoyancy while a loop initialized with @xmath224 is strongly influenced by convective motions .
similar values are found here .
figure [ figure_density ] shows the evolution of the density deficit inside the loop compared to its surroundings , for case 1 ( dotted line ) and case 3 ( plain line ) .
this density deficit represents the buoyancy of the structure and thus its capacity to reach the top of our computational domain .
both loops started with approximately the same density deficit , as seen at the beginning of the evolution .
the two cases first follow the same amount of buoyancy loss during about half a day and then the two cases start to have different behaviors .
indeed , in case 1 , the loops continues to smoothly lose its density deficit while the simulation proceeds and follows the exponential decay of a typical solution of diffusion equation .
after @xmath217 days , the density fluctuations in the bulk of the convection zone are of the same order as the density deficit in the loop and the magnetic structure stops being buoyant .
we also note that convective motions act on the magnetic structure to create smaller scale regions which will thus diffuse faster and lead to a smaller field strength and thus smaller buoyancy . at @xmath146 days ,
the loop has still not reached the top of the domain and the only reason why we will get some emerging magnetic field in this case is that upflows will drag the loop upward .
the coherent bipolar structure will then be mostly lost when the magnetic structures will emerge at @xmath184 , as we will see in the following section . in case 3 on the contrary ,
the evolution of the density deficit is quite different .
the major trend is of course a decrease of the buoyancy of the structure but this decrease does not follow a decaying exponential as we had in the previous case . here , strong currents generated by the higher field strength ( initially @xmath215 ) will play a role in the energy balance of the loop and the diffusion of entropy will be less efficient in this case . as we see on the figure ,
the buoyancy of the loop of case 3 always stay higher than in case 1 . as a consequence , at @xmath146 days , a significant part of the loop has emerged at the surface as a coherent bipolar structure .
the loop thus keeps both its coherence and its buoyancy for a longer time and the structure of the emerging regions will tend to be less influenced by the convective motions .
as stated in the introduction , one of the main questions we wish to answer in this work is the following : `` what is the influence of convective motions and large - scale flows on the structure of our emerging magnetic field ? '' .
although our top boundary condition is still at @xmath225 below the surface and that density and pressure drops may well cause large modifications of the observed magnetic field , we can get some insight on what type of active regions may be created at the solar surface from our numerical simulations .
in this section , we focus on the _ convection dominated regime _ of case 1twp ( and 1twn ) , where the emerging magnetic field will be strongly affected by the convective motions .
figure [ figure_shsl1 ] shows four snapshots of the emerging region zoomed in latitude and longitude , at @xmath193 .
the radial velocity ( colors ) represents the convective motions ( blue downdrafts and yellow upflows ) and we superimpose the emerging radial field ( red contours indicating positive polarity ) . at the beginning of emergence at about @xmath226 days
, a well - defined bipolar structure can be identified , at the latitude of @xmath133 , indicating that the rise was indeed radial and not parallel to the rotation axis .
we note already on this snapshot that the strong downflow located at the latitude of @xmath133 and the longitude of @xmath227 strongly interacts with the negative polarity and is responsible for its cashew - nut - like shape .
as the emergence proceeds , the well - identified bipolar structure is replaced by a much more complex field structure .
the negative polarity gets squeezed into the previous strong downflow , while regions of positive polarity start to emerge all around .
the concentration of magnetic field in the downflow lanes is even more visible on the third snapshot at @xmath228 days where the field concentration seen on the previous panel continues to accumulate in the downdraft while a new emerging bipolar structure appears at the longitude of @xmath171 where a large upflow was already present and responsible for this new emergence .
we then note that this new region gets quickly advected toward the boundary of this large upflow to , in turn , get concentrated into the neighboring downdrafts , leading to an evolution very close to what is observed in simulations of magnetoconvection @xcite . in this case ,
convective motions were thus very weakly affected by the emerging magnetic field and completely dominates its evolution .
the evolution shown in the previous figure is quite different when cases with stronger buoyancy and a stronger initial field strength are considered .
we now turn to investigate the structure of the magnetic field emerging in cases 2twp and 4twp which initially possess the same buoyancy but different field strengths ( @xmath215 for case 2twp and @xmath229 for case 4twp ) .
these values respectively correspond to @xmath216 and @xmath230 times the equipartition field strength .
the results are shown in figure [ figure_shsl2 ] .
four snapshots are shown , two for each case , at approximately the same times , @xmath226 days and @xmath231days .
case 2twp is interesting since it is an intermediate case between _ convection dominated _ and _ magnetic field dominated _ evolution .
indeed , at the beginning of emergence , a bipolar structure emerges at about @xmath232 as expected ( and as found in the isentropic cases ) but another region , which is in fact another part of the loop , also emerges at the longitude of @xmath227 .
this is due to the strong downflow lane at about @xmath160 of longitude and its neighboring upflow which will have the effect of dragging the loop upwards in the upflow and pinning it down in the downflow ( those flows are also visible in figure [ figure_eqslc ] ) . as a consequence ,
two separate regions seem to emerge when they are part of the same @xmath0-loop but strongly influenced by convection .
as the simulation evolves and emergence proceeds , we recover that the magnetic field gets advected in the downflow lanes at the largest longitudes but shapes as a typical active region at around @xmath171 .
indeed , the tongue - shape appears here , although modified by the convective motions .
this typical case , intermediate between a _ convection dominated _ and _ magnetic field dominated _ evolution , is also interesting since it directly relates to observations of emerging regions .
indeed , the regions of strong magnetic field , in the center of the active regions , are only marginally affected by the convective motions , as opposed to the surrounding plage regions whose structure is affected by the supergranular scales of convection .
we could thus consider that the parameters of case 2 enable us to simulate an emerging region with characteristics typical of an observed active region , in particular its field strength in each polarity , its competition with convective motions , its tongues or its tilt angle .
if we now look at case 4twp where the initial magnetic field was stronger , the emergence is much more similar to what we had in the isentropic cases . the orientation of the unique ( unlike case 2 ) structure is first north - south to become more and more east - west , the tongues are easily identified , a strong upflow is created by the loop itself and dominates its evolution during the emergence process .
we observe in figure [ figure_shsl2 ] that the apex of the loop emerges at the longitude of @xmath232 and that no radial field is visible at @xmath227 in case 4 , contrary to case 2 .
this is due to the fact that in case 2 , the advection of the magnetic structure by the convective motions significantly modifies the shape of the loop and thus the asymmetry between the trailing and leading legs . in case 4 ,
the evolution is again closer to the isentropic case and convection has little influence on the loop and is thus not strong enough to produce an emerging region within the upflow located at @xmath227 . to conclude on these cases ,
although case 2 is more influenced by convection , a clear impact of the emerging structure on the velocity field is visible in both cases , especially at @xmath231 days where a strong upflow is created at the very location of emergence . in this last section ,
we investigate the effect of changing the sign of the twist of the initial magnetic loop , as well as modifying its latitude of introduction .
we saw in the isentropic cases that the sign of the twist would influence the tilt angle of the emerging region but that a tilt compatible with observations could still be found with an initial left - handed twist .
we now address the question of the possibility to get tilt angles even more compatible with observations when the magnetic loop is embedded in a convective environment . before specifically studying the tilt induced by the twisted structures , we investigate the impact of a negative initial twist on the intensity of the emerging radial field .
indeed , it was already found in the isentropic cases that the strength of the emerging polarities and the positive and negative fluxes had a different evolution when a left - handed or right - handed twist was considered .
it is still the case when the loops are introduced in a convective shell .
figure [ figure_br ] presents the evolution of the radial field of the leading and trailing polarities at the time of and after emergence for cases 1twp , 1twn , 3twp and 3twn .
similar pictures for cases 1twp and 1twn were shown in figure [ figure_shsl3 ] for the isentropic simulations .
we can thus first compare the convective and isentropic situations for case 1 .
the first thing we note and which is true both for the right - handed and left - handed twist , is that the loop in the isentropic case had a quicker rise than in the convective case .
similar results were found for uniformly buoyant flux tubes in @xcite .
indeed , for example at @xmath217 days for the isentropic cases , the typical field strength already reached @xmath233 whereas it barely gets to @xmath234 in the convective cases .
this is related to the explanations of section [ sect_bulk ] where the buoyancy loss was shown and thought to be partially due to the convective motions creating smaller scales and thus stronger dissipation .
another striking feature is the strong increase of negative @xmath175 in case 1twp ( panel ( @xmath51 ) ) at @xmath219 days and of positive @xmath175 in case 1twn ( panel ( @xmath182 ) ) at about the same time .
this is mainly due to the fact that these cases are dominated by convective motions . indeed , as we saw in figure[figure_shsl1 ] , the negative polarity in case 1twp between @xmath235 and @xmath236 days gets strongly squeezed by the convective motions in the narrow downflow lanes .
this compression of magnetic field tends to produce stronger concentration of fields , explaining the strong increase of negative @xmath175 in case 1twp and then its sharp decrease due to fast dissipation of small scales .
a similar situation happens in case 1twn for the opposite dominating polarity . in cases 3twp and 3twn ( panels ( @xmath180 ) and ( @xmath181 ) ) where the initial magnetic field is twice as large , such sharp increase and decrease of the two polarities are not visible since the loop emerges as a global magnetic structure and is much less influenced by convective motions .
it is worth noticing that we may have a possible way here to explain the formation of complex active regions from where powerful ( x - class ) flares could originate . indeed ,
active regions with a complex topology could be created by the turbulent convective motions acting on moderately strong flux tubes ( of approximately @xmath63 at the base of the cz ) . by the time they reach the top of the convection zone ,
the flux tubes have been strongly deformed and the subsequent emerging radial field consists in several patches of opposite polarities , as seen in figure [ figure_shsl1 ] .
we could argue that this complex active region , whose progenitor is a relatively weak flux tube , would not be able to produce very energetic flares .
however , when panel ( @xmath51 ) of figure [ figure_br ] is considered , we conclude that the converging motions at the sub - photospheric level are able to re - concentrate the magnetic structure in the strong downdrafts , in order to locally produce a strong increase of the radial field .
in particular , we note that the maximum positive radial field is about twice as large as the negative @xmath175 , which was not the case in the isentropic case where the re - concentration by convective motions could not occur .
this scenario may thus explain the injection of energy in the corona by convective motions acting on complex active regions originating from initially weak flux tubes .
as far as the tilt angle is concerned in these simulations , the influence of the initial twist , which is not included in the thin - flux tube calculations , is again crucial .
if we follow the position in latitude of the positive trailing and negative leading polarities ( not shown here ) , we find that , at the beginning of emergence , the leading polarity is located at higher latitude than the leading one , unlike in solar observations .
however , in case 1twn , the clockwise rotation due to the coriolis force acting on the legs of the loop tends to reverse this situation about @xmath134 days after emergence , implying a leading polarity at slightly lower latitude than the trailing one and thus to a tilt of about @xmath237 , in agreement with joy s law .
this is similar to what was found in the isentropic case 1twn .
case 3twn does not follow the same evolution .
indeed , the orientation at emergence is kept even after @xmath121 days after emergence before the magnetic field is swept towards the downflow lanes and dissipates .
again , we need a loop that is sufficiently influenced by the coriolis force to get a tilt in agreement with observations with an initial left - handed twist believed to dominate in the northern hemisphere .
but for the convective cases 2twp15 , 2twp , and 2twp60.,width=283 ] of course , not only will the initial twist play a role in the tilt of the emerging region but also the latitude of introduction and emergence , as already pointed out in the isentropic cases .
we thus computed extra cases as in the isentropic simulations , cases 2twp15 and 2twp60 .
we did not choose the same initial parameters as in the isentropic cases to study the effect of latitude , since case 1 was computed in the isentropic case and it is believed here to be influenced too much by convection to produce a significant trend and thus to allow drawing any conclusions on the tilt at the time of emergence at @xmath238 .
case 2twp is the most interesting since it is still influenced by convection ( as we saw in the previous section ) without completely loosing its characteristic bipolar structure when the loop emerges at the top of our domain .
the results of these extra calculations are shown in figure [ figure_convlat ] .
both the emerging radial magnetic field and a measure of the tilt angle are represented , similar to figure [ figure_shsl4 ] .
we argue that some effects of the differential rotation are visible here .
first , the rise velocities are different from one case to another , like the isentropic simulations but the differences are even more pronounced here . to be more precise ,
the characteristic rise time for the loop initially located at a latitude of @xmath99 is around @xmath218 days , when it is of more than @xmath239 days for the case at latitude @xmath240 .
the figures represent the emerging region about @xmath241 to @xmath242 days after the first signs of emergence at @xmath193 .
this slow down of the loops located at lower latitudes is related to the rotation rate which is stronger at lower latitudes .
indeed , as shown already in @xcite , the modified buoyancy force in the presence of rotation is proportional to @xmath243 and the rise velocity of the loop will thus be less at low latitudes and high rotation rates , which is typically the situation for case 2twp15 represented here in the first panel of figure [ figure_convlat ] . a second effect of the differential rotation which can be noted concerns the longitude of emergence . in these differentially rotating cases ,
the loop which was introduced at higher latitudes has drifted in longitude by about @xmath133 ( it was introduced at @xmath160 and emerges at @xmath244 ) when the loop in case2twp15 has only drifted by @xmath237 .
this is quite different from what we had in the isentropic cases where the drift in longitude was more important for the case at low latitudes .
this was due to the stronger asymmetry between the two legs of the loop which appeared in the low latitude case . here
, the different drift values could well be due again to the differential rotation .
indeed , since the higher ( lower ) latitudes rotate retrograde ( prograde ) compared to the rotating frame , it is not surprising that the loop at high latitudes will be globally shifted at lower longitudes compared to the low latitude cases .
this is thus a clear effect here of the large - scale differential rotation acting on our rising @xmath0-loops . finally , the idea here was to compare the tilt angles we could get in these convective cases and compare those values to observations .
it is more complicated here to identify well - defined concentrations of opposite polarities , since the convective motions have influenced the loops during their rise to sometimes produce several concentrations of radial magnetic field of both polarities .
however , if we consider that the tilt angle is defined as the angle between the east - west direction and the line joining the maximum of both polarities , we are still able to measure a tilt in those 3 cases . we find a value of about @xmath245 for case2twp15 , @xmath246 for case2twp and @xmath168 for case2twp60 .
the trend is thus still to increase with increasing latitude but it is less obvious than in the isentropic cases .
we note here that the tilt angle distribution in the simulations of @xcite was also much more spread in the convective cases than in the isentropic ones , which is consistent with our simulations .
this is an indication here that it is not only the rotation and the associated coriolis force acting on the loop which will determine the tilt angle observed at the solar surface but horizontal buffeting by convection could also strongly modify this measurement , even if the trend compatible with joy s law is still found in those cases .
in this work , we presented the dynamical evolution of twisted @xmath0-loops embedded into a convectively stable and unstable unmagnetized environment , to mimic the behavior of a strong toroidal magnetic field generated by dynamo action at the base of the solar convection zone and rising toward the photosphere . those simulations allow to take into account the full effects of sphericity , mean flows and convective motions on the dynamical evolution of a flux rope with a finite radius and thus contains much more physics than previous thin flux tube calculations .
the first part of this work was dedicated to establish the important parameters to get a coherent radial rise in the isentropic case .
we find that a loop with an initial magnetic field of @xmath63 and with an initial twist of less than 1 turn along the loop extension ( @xmath100 here ) is able to rise and emerge at the top of our domain to create bipolar regions .
asymmetries develop between the leading and trailing legs of the loop due to the combined effect of angular momentum conservation and lorentz forces and it is found that this asymmetry is strongly reduced when magnetic structures as strong as @xmath247 are considered .
the emerging regions obtained in these isentropic cases have several properties in agreement with observations .
in particular , well - identified tongues develop during emergence , mainly due to the initial twist of the magnetic field lines , also seen in observations ( e.g. * ? ? ?
for the first time , we point the appearance during flux emergence of sharp ring - shaped magnetic structures around the bipolar region , that we call `` magnetic necklaces '' and which are due to the creation of vorticity at the loop periphery .
they are also very often seen in observations of large - scale flux emergence ( especially in @xcite with mdi data ) and are very similar in shape to the ones we get in our simulated active regions .
moreover , we show here that it is possible to get a case where the initial left - handed twist is strong enough to maintain the coherence of the loop and where the tilt angle of the emerging radial field agrees with joy s law ( in contrast to what is found in @xcite ) .
to do so , the rise time needs to be long enough so that the coriolis force acts sufficiently on the loop to rotate the two opposite polarities clockwise and then produce a final tilt angle of the correct sign .
we note that if much thinner flux tubes were considered , the various parameters quoted here to get bipolar regions with properties in agreement with observations would probably have to be revised .
the results do not change significantly if a tube four times smaller is introduced as long as the magnetic diffusivity is also reduced to keep a rise time small compared to the diffusion time of the magnetic structure ( see * ? ? ? * for a discussion on the effect of diffusion on this type of simulations ) . however ,
if much thinner tubes are introduced ( say 10 to 20 times smaller as can be expected in the real sun ) , it is likely that the results may differ . in particular
, the drag force coming from the pressure difference between the up and down stream sides of the turbulent wake behind the rising loop may be much bigger when the loop radius is reduced .
this was shown in @xcite for incompressible flows passing a rigid cylinder under large reynolds number conditions and shown to be still true for two - dimensional simulations of buoyant loops @xcite .
this will in turn modify the rise velocity and the competition between magnetic fields and convective motions .
for example , a smaller rise velocity leaves more time for the coriolis force to act but at the same time , an increased drag directly opposes the tilting motion , resulting in a net decrease of the tilt angle of loops of smaller radii , as shown by @xcite .
this work would then benefit from studying the effect of loops of smaller radii but this question was not addressed here .
@xcite also find that the wake structure is modified by the presence of a background dynamo field , possibly due to continuous field reconnection as the tube rises . in the convective cases ,
the rise velocity and characteristics of emerging regions are strongly affected by the convective motions when loops of less than @xmath139 are considered . however , emerging regions with the correct orientation and a dominant leading polarity are still found in these simulations where the full effects of convection , rotation and sphericity are taken into account . making direct comparison with observed photospheric field intensity in active regions is not straightforward since the scaling of the magnetic field strength with density ( which is linear in our cases ) may change drastically when reaching the uppermost layers of the convection zone @xcite .
however , local helioseismology techniques seem to start to provide us with constraints on the field intensity deeper down @xcite , much closer to the top of our computational domain .
we may well soon be able to relate much more directly the results of our three - dimensional simulations to actual observations . if the origin of the tilt angle of active regions has not been completely related to one particular property of the emerging loops yet , it is however undeniable that the initial twist of the field lines is an important ingredient
. it is thus necessary to understand the generation not only of buoyant magnetic field at the base of the convection zone but of _ twisted _ buoyant structures .
it is believed that such twisted loops could be created by the advection of poloidal field lines by the initially rising strong toroidal field subject to magnetic buoyancy instabilities @xcite .
full mhd simulations of dynamo - generated toroidal field becoming buoyant , gaining some twist and being able to rise coherently to the top of the convection zone are in progress @xcite .
we moreover note that such simulations are able to mimic the dynamical evolution of a magnetic structure from the base of the convection zone to about @xmath103 mm under the photosphere and that emergence higher up is not possible in these global anelastic simulations .
contact with three - diemnsional compressible local simulations of emerging flux in the solar atmosphere @xcite should now be considered to have a full picture of the mhd evolution of large - scale magnetic regions .
when convection is considered , the emerging radial field may lose the typical characteristics of a well - defined bipolar structure that we get in the isentropic cases .
indeed , in case 1twp where the evolution was dominated by the downflows and upflows of the surrounding convection , the emerging structure is very complex and regions of mixed - polarity are self - consistently formed
. the large variety of large - scale emerging regions at the solar photosphere may then well be the results of magnetic @xmath0-loops significantly processed by convection , even if robust statistical properties such as tilt angle or asymmetry between polarities indicate that these loops are still able to keep their coherence during their rise .
another possibility leading to the emergence of complex active regions in the sun could be that such buoyant structures might interact in the bulk of the convection zone . depending on
which kind of reconnection could occur between the @xmath0-loops , either complex region with mixed polarity and mixed helicity could emerge ( such as the one observed by @xcite ) , or the magnetic field could completely change its trajectory , leading to no emergence at all ( see @xcite for a discussion of reconnection between twisted flux tubes ) .
this is the analysis we intend to focus on in the near future .
further no background dynamo field was present in the simulation .
@xcite have started to address the influence of such a dynamo field on a uniformly buoyant magnetic rope .
we intend to carry on this analysis for @xmath0-loops .
we thank the organizers of the flux emergence workshops held in st andrews in 2007 june , in kyoto in 2008 october , and in berkeley in 2011 august .
we also acknowledge financial support by the erc through grant 207430 stars2 and fundings by the programme national soleil terre ( insu / pnst ) .
we also appreciate the genci supercomputing centers for granting access to their infrastructures through project 1623 .
abbett , w. p. , fisher , g. h. , & fan , y. 2000 , , 540 , 548 acheson , d.j .
1979 , solar physics , 62 , 23 archontis , v. , & trk , t. 2008 , , 492 , 35 archontis , v. , & hood , a. w. 2010 , , a62 aulanier , g. , trk , t. , dmoulin , p. , & deluca , e.e .
2010 , , 708 , 314 batchelor , g.k .
1967 , _ an introduction to fluid dynamics _ , cambridge university press , cambridge , u.k .
braun , d. c. 2012 , science , 336 , 296 browning , m. k. , miesch , m. s. , brun , a. s. , & toomre , j. 2006 , , 648 , l157 brun , a. s. , antia , h. m. , chitre , s. m. , & zahn , j .-
2002 , , 391 , 725 brun , a. s. 2004 , solar physics , 220 , 333 brun , a.s . ,
miesch , s.m . , & toomre , j. 2004 , apj , 614 , 1073 caligari , p. , moreno - insertis , f. , & schssler , m. 1995 , apj , 441 , 886 caligari , p. , schssler , m. , & moreno - insertis , f. 1998 , apj , 502 , 481 cattaneo , f. , & hughes , d.w .
1988 , j. fluid mech . , 196 , 323 chae , j. , & moon , y .-
2005 , , 629 , 1110 chandra , r. , pariat , e. , schmieder , b. , mandrini , c.h .
, & uddin , w. 2010 , solar physics , 261 , 127 cheung , m. c. m. , rempel , m. , title , a. m. , & schssler , m. 2010 , , 720 , 233 charbonneau , p. 2005 , living rev .
solar phys . , 2 , 2 choudhuri , a. , & gilman , p. 1987 , apj , 316 , 788 clune , t. l. , elliott , j. r. , glatzmaier , g. a. , miesch , m. s. , & toomre , j. 1999 , parallel comput . , 25 , 361 dmoulin , p. , mandrini , c.h . , van driel - gesztelyi , l. , et al . 2002 , , 382 , 650 dsilva , s. , & choudhuri , a. r. 1993 , , 272 , 621 emonet , t. , & moreno - insertis , f. 1998 , apj , 492 , 804 fan , y. 2001 , , 554 , 1 fan , y. 2004 , living rev .
, 1 , 111 fan , y. 2008 , , 676 , 680 favier , b. , jouve , l. , edmunds , w. , silvers , l.j . , & proctor , m.r.e . 2012 , , 426 , 3349 gilman , p. a. , & glatzmaier , g. a. 1981 , , 45 , 335 hale , g.e . ,
ellerman , f. , nicholson , s.b . , & joy , a.h .
1919 , , 49 , 153 ilonidis , s. , zhao , j. , & kosovichev , a. 2011 , science , 333 , 993 ilonidis , s. , zhao , j. , & kosovichev , a. 2012 , science , 336 , 296 jouve , l. , & brun , a.s .
2007 , astron .
notes , 328 , 1104 jouve , l. , & brun , a.s .
2009 , , 701 , 1300 kosovichev , a. g. , & stenflo , j. o. 2008 , , 688 , l115 leka , k.d . ,
canfield , r.c .
, mcclymont , a.n .
, & van driel - gesztelyi , l. 1996 , , 462 , d547 linton , m. , & antiochos , s.k .
2005 , , 625 , 506 liu , j. , & zhang , h. 2006 , sol .
phys . , 234 , 21 longcope , d. w. , fisher , g. h. , & arendt , s. 1996 , , 464 , 999 lopez fuentes , m.c . , dmoulin , p. , mandrini , c.h . ,
& van driel - gesztelyi , l. 2000 , , 544 , 540 luoni , m.l . , dmoulin , p. , mandrini , c.h . , &
van driel - gesztelyi , l. 2011 , sol .
phys . , 270 , 45 magara , t. 2004 , , 605 , 480 manchester , w. , iv , gombosi , t. , dezeeuw , d. , & fan , y. 2004 , , 610 , 588 martnez - sykora , j. , hansteen , v. , & carlsson , m. 2008 , , 679 , 871 matthews , p.c . , hughes , d.w . , & proctor , m.r.e . 1995 , , 448 , 938 miesch , m. s. 2005 , living rev .
solar phys . , 2 , 1 miesch , m. s. , brun a. s. , & toomre j. 2006 , , 641 , 618 miesch , m. s. , elliott , j. r. , toomre , j. , et al .
2000 , , 532 , 593 moffatt , h. k. 1978 , _ magnetic field generation in electrically conducting fluids _
( cambridge univ .
press ) , 353 moreno - insertis , f. , schssler , m. , & caligari , p. 1994
phys . , 153 , 449 moreno - insertis , f. , schssler , m. , & ferriz mas , a. 1992 , a&a , 264 , 686 murray , m. j. , hood , a. w. , moreno - insertis , f. , galsgaard , k. , & archontis , v. 2006 , , 460 , 909 nelson , n.j . , brown , b.p . , brun , a.s . ,
miesch , m.s . , & toomre , j. 2011 , , 739 , 38 nelson , n.j . ,
brown , b.p . ,
brun , a.s . ,
miesch , m.s . , & toomre , j. 2012 , , accepted ossendrijver , m. 2003 , a&ar , 11 , 287 otsuji , k. , kitai , r. , ichimoto , k. , & shibata , k. 2011 , pasj , 63 , 1047 parker , e. n. 1955 , , 122 , 293 parker , e. n. 1993 , , 408 , 707 pariat , e. , aulanier , g. , schmieder , b. , et al .
2004 , , 614 , 1099 pinto , r. , & brun , a.s .
2012 , , under review pomarde d. , & brun , a.s .
2010 , in asp conf .
434 , astronomical data analysis software and systems xix , ed .
y. mizumoto , k .-
morita , & m. ohishi ( san francisco , ca : asp ) , 378 sanchez almeida , j. 1998 , in asp conf .
155 , in three dimensional structure of solar active regions , ed .
alissandrakis & b. schmieder,(san francisco , ca : asp ) , 54 schrijver , c. j. , & harvey , k. l. 1994 , , 150 , 1 schssler , m. , caligari , p. , ferriz - mas , a. , & moreno - insertis , f. 1994 , , 281 , 69 spruit , h.c . 1981 , a&a , 98 , 155 trk , t. , & kliem , b. 2005 , , 630 , l97 van driel - gesztelyi , l. 2002 , in proc .
magnetic coupling of the solar atmosphere euroconference , solmag 2002 , ed .
sawaya - lacoste ( esa sp-505 ; nordjwik : esa ) , 113 van driel - gesztelyi , l. , dmoulin , p. , mandrini , c.h . ,
harra , l. , & klimchuk , j.a .
2003 , , 586 , 579 wang , y .- m . , &
sheeley n.r.,jr .
1989 , sol .
phys . , 124 , 81 wang , y .- m . , & sheeley n.r.,jr .
1991 , , 375 , 761 weber , m.a . , fan , y. , & miesch , m.s .
2011 , , 741 , 11 weiss , n.o . ,
brownjohn , d.p . ,
matthews , p.c . , & proctor , m.r.e .
1996 , , 283 , 1153 wong , v. c. , & lilly , d. k. 1994 , physics of fluids , 6 , 1016 zwann , c. 1987 , ara&a , 25 , 83 | we present three - dimensional numerical simulations of a magnetic loop evolving in either a convectively stable or unstable rotating shell .
the magnetic loop is introduced in the shell in such a way that it is buoyant only in a certain portion in longitude , thus creating an @xmath0-loop . due to the action of magnetic buoyancy ,
the loop rises and develops asymmetries between its leading and following legs , creating emerging bipolar regions whose characteristics are similar to the ones of observed spots at the solar surface . in particular , we self - consistently reproduce the creation of tongues around the spot polarities , which can be strongly affected by convection .
we moreover emphasize the presence of ring - shaped magnetic structures around our simulated emerging regions , which we call `` magnetic necklace '' and which were seen in a number of observations without being reported as of today .
we show that those necklaces are markers of vorticity generation at the periphery and below the rising magnetic loop .
we also find that the asymmetry between the two legs of the loop is crucially dependent on the initial magnetic field strength .
the tilt angle of the emerging regions is also studied in the stable and unstable cases and seems to be affected both by the convective motions and the presence of a differential rotation in the convective cases . |
after the discovery of superconductivity in lafeaso @xcite , the finding of iron - chalcogenide superconductor @xmath7-fese@xmath8 @xcite has stimulated another intensive studies on the iron - based superconductors .
although the iron - chalcogenides and iron - pnictides have similar crystal and electronic structures , they show many differences either in superconductivity or in magnetism .
for instance , the former has lower @xmath9 ( @xmath10 ) @xcite and a larger magnetic moment of the fe ion in fete@xmath11se@xmath8 ( @xmath12 ) , while the latter , _ e.g. _ la-1111 , has much higher @xmath13 @xcite but has a smaller magnetic moment ( @xmath14 ) for the fe ion@xcite .
even the magnetic orders are different , one is bi - collinear and the other is collinear .
recently , k , cs or rb intercalated fese superconductors a@xmath0fe@xmath15se@xmath16 are found .
it is shown that @xmath9 of this system can be enhanced above @xmath17 @xcite . in this system , as the atomic ratio of fe : se is not 1:1 , which used to be in iron - chalcogenides , iron deficiency is produced . as a result , in addition to the enhancement of @xmath9 , a @xmath18 iron vacancy ordered pattern illustrated in fig .
[ lattice_245 ] is formed when the composition is close to k@xmath0fe@xmath19se@xmath2 @xcite .
the vacancy ordering is followed by the magnetic transition at lower temperature @xmath20 to the block antiferromagnetic ( afm ) phase with a big moment 3.31@xmath21 @xcite .
although the afm state is observed in superconducting k@xmath22fe@xmath1se@xmath16 , evidence of nanoscale phase separation from x - ray diffraction is reported @xcite .
previous theoretical works @xcite on superconductivity in the a@xmath0fe@xmath1se@xmath2 system were based on the band structure of kfe@xmath2se@xmath2 in which the hole pocket around @xmath23 is absent and only electron pockets are present at m. such fermi surface tomography was supported by arpes @xcite .
however , since there is no experimental evidence @xcite showing the existence of the @xmath24 pattern in kfe@xmath2se@xmath2 , the validity of these approaches is questionable .
in fact , because of the @xmath25 pattern , the symmetry group changes from _ i4/mmm _ to _ i4/m _ and both the unit cell and the brillouin zone ( bz ) change as well . the underlying band structure should be very different .
indeed , the first - principles calculations of the @xmath24 vacancy ordered lattice structure @xcite indicate that a hole pocket at @xmath26 appears in the nonmagnetic state and as expected it loses the reflection symmetry in the _ x - y _ plane .
the presence of a hole pocket at @xmath23 point indicates that the physics that drives superconductivity could be very different .
it thus calls for a close examination based on an appropriate hamiltonian to model the ternary iron selenide k@xmath0fe@xmath1se@xmath2 . in this paper
, we construct a two - dimensional tight - binging model with three @xmath27 orbitals for the k@xmath22fe@xmath1se@xmath2 system with the @xmath18 vacancy ordered pattern being explicitly included . based on the general tight - binding model @xmath28 with symmetry imposed by the vacancy order
, we fit the dispersion relation of @xmath28 to that of the non - magnetic state from the first - principles calculations of ref .
two hole pockets and two electron pockets emerges in the fitted tight - binding model .
the constructed band model , when combined with generalized hubbard interactions , yields a spin susceptibility that shows a large peak around the @xmath5-vector ( @xmath6 ) for the block - checkerboard afm state .
furthermore , competition of different magnetic states is found but the block - checkerboard afm state gets enhanced with larger hund s rule coupling and wins out at the end .
the implication of our results to the mechanism of superconductivity is discussed . in particular
, we argue that the spin fluctuations at @xmath5=(@xmath6 ) suggest that in analogy to spin - fluctuations in high @xmath9 cuprates , the interblock fluctuations of spins might play an important role in the mechanism of superconductivity occurring in this system .
we start by considering the vacancy - ordered structure with right - handed chirality @xcite as shown in fig .
[ lattice_245 ] . in this ordered state ,
the unit cell includes four iron atoms ( without considering se ) , which we denote as @xmath29 , @xmath30 , @xmath31 , and @xmath32 and three * * * * @xmath27 * * * * orbitals ( @xmath33 , @xmath34 , and @xmath35 ) are considered in each iron
. therefore we have 12 species of electrons in a unit cell .
we will suppress spin indices and denote the electron operators collectively as a vector by @xmath36 with @xmath37 , where @xmath38 denotes the annihilation operator of electron for orbital @xmath39 at site @xmath40 with @xmath41 , @xmath42 , @xmath43 standing for @xmath44 , @xmath45 , @xmath35 , respectively .
we will use @xmath46 and @xmath47 as the coordinates of this system and @xmath48and @xmath49 as the nearest fe - fe directions . in previous works , the dft band structure of the parent compound kfe@xmath2se@xmath16
in which only electron pockets appear at m tamaier , fawang , yzhou , tdas1,mazin is employed .
however , the parent compound kfe@xmath2se@xmath2 does not have the same hopping parameters as those in the vacancy - ordered k@xmath0fe@xmath1se@xmath2 .
for example , there is no hopping between fe atoms and vacancies .
here we shall strictly enforce the @xmath3 vacancy - ordered structure and construct a tight - binding model with nearest neighbor ( nn ) and next - nearest neighbor ( nnn ) fe - fe hoppings , which will be classified as intra- and inter - cell ones .
as the vacancy ordering appears , the reflection symmetry is lost but the four - fold - rotational symmetry is left intact . the system is invariant under 90@xmath50 * * * * rotations around the center of a unit cell , which is the position of se ( if the vacancy position is taken as the rotation center , the rotation has to be followed by a @xmath51 operation : the reflection @xmath52 ) .
a general hopping hamiltonian @xmath28 with vacancy order being included can be written down by imposing the 90@xmath53 right - handed rotation symmetry : @xmath54 , @xmath55 , @xmath56 accompanied with @xmath57 , @xmath58 , @xmath59 , @xmath60 . due to its massive form
, the general form of @xmath28 is relayed to the appendix .
after fourier transformation , in the momentum @xmath61 space , it takes the form @xmath62where @xmath63 with @xmath64 and @xmath65 is a @xmath66 matrix .
detailed characterization of all hopping parameters are tabulated in table [ intra_t ] and table inter_t in appendix .
these parameters are obtained by fitting energy dispersions ( in the folded bz ) to the results of x. w. yan _ et al . _
@xcite obtained by the generalized gradient approximation ( gga ) in which the main features are four hole pockets at @xmath23 and four electron pockets at x in the nonmagnetic state .
our fitting gives two hole and two electron pockets , which capture basic features of this system .
[ ek ] shows our fitting results in the unfolded bz ( ( a ) and ( c ) ) , and in the folded bz ( ( b ) and ( d ) ) . in the folded bz ,
there are two hole pockets around ( @xmath67 ) and two electron pockets around ( @xmath68 ) ; in the unfolded coordinate , one hole pocket will move to ( @xmath6 ) and electron pockets to @xmath69(@xmath70 ) .
a. ricci , n. poccia , b. joseph , g. arrighetti , l. barba , j. plaisier , g. campi , y. mizuguchi , h. takeya , y. takano , n. lal saini , and a. bianconi , supercond .
sci . tech .
* 24 * , 082002 ( 2011 ) ; a. ricci , n. poccia , g. campi , b. joseph , g. arrighetti , l. barba , m. reynolds , m. burghammer , h. takeya , y. mizuguchi , y. takano , m. colapietro , n. l. saini , a. bianconi , e - print arxiv:1107.0412 .
y. zhang , l. x. yang , m. xu , z. r. ye , f. chen , c. he , h. c. xu , j. jiang , b. p. xie , j. j. ying , x. f. wang , x. h. chen , j. p. hu , m. matsunami , s. kimura , and d. l. feng , nature mate . * 10 * , 273 ( 2011 ) .
d. mou , s. liu , x. jia , j. he , y. peng , l. zhao , l. yu , g. liu , s. he , x. dong , j. zhang , h. wang , c. dong , m. fang , x. wang , q. peng , z. wang , s. zhang , f. yang , z. xu , c. chen , and x. j. zhou , phys .
. lett . * 106 * , 107001 ( 2011 ) .
l. zhao , d. mou , s. liu , x. jia , j. he , y. peng , l. yu , x. liu , g. liu , s. he , x. dong , j. zhang , j. b. he , d. m. wang , g. f. chen , j. g. guo , x. l. chen , x. wang , q. peng , z. wang , s. zhang , f. yang , z. xu , c. chen , and x. j. zhou , phys .
b * 83 * , 140508(r ) ( 2011 ) . | we investigate the electronic structure of the ternary iron selenide k@xmath0fe@xmath1se@xmath2 by considering the spatial symmetry of the @xmath3 vacancy ordered structure .
based on three orbitals of @xmath4 , which are believed to play major physics in iron - based superconductors , an effective two - dimensional tight binding hamiltonian is constructed with the vacancy ordered structure being explicitly included . it is shown that the constructed band model , when combined with generalized hubbard interactions , yields a spin susceptibility which exhibits both the block - checkerboard antiferromagnetism instability and the stripe antiferromagnetism instability .
in particular , for large hund s rule couplings , the block - checkerboard antiferromagnetism wins over the stripe antiferromagnetism , in agreement with the observation in experiments .
we argue that such a model with correct symmetry and fermi surface structures should be the starting point to model k@xmath0fe@xmath1se@xmath2 .
the spin fluctuations at @xmath5=(@xmath6 ) suggest that interblock fluctuations of spins might play an important role in the mechanism of superconductivity occurring in this system . |
symbiotic stars are believed to be wide binary systems consisting of a hot component , usually a white dwarf , and a mass losing giant ( e.g. @xcite , @xcite ) . because of the size of the giant component , their orbital periods range from a few hundred days to several decades ( e.g. @xcite ) .
symbiotic stars exhibit prominent emission lines indicative of a wide range of ionization including [ sii ] and o vi .
some fraction of the slow stellar wind from the giant component is accreted to the hot component , which leads to various activities including occasional eruptions and strong emission lines ( e.g. @xcite , @xcite ) . @xcite
performed _ smoothed particle hydrodynamical _
computations for binary systems of a mass losing giant and a white dwarf to show that an accretion disk may be formed through accretion of the stellar wind .
however , it is still an unresolved issue whether an accretion disk is formed in general in symbiotic stars ( e.g. @xcite , @xcite ) .
symbiotic stars are classified into s type and d type , where the latter exhibit an ir excess indicating the existence of a warm dust component surrounding the binary system .
it appears that d - type symbiotic stars possess multiple dust shells with temperatures of @xmath6 and @xmath7 ( @xcite ) .
many symbiotic stars are also x - ray emitters exhibiting a large range of hardness .
@xcite classified x - ray emitting symbiotic stars into types @xmath8 and @xmath9 depending on the hardness with the @xmath10 type being supersoft x - ray sources and the @xmath9 type being highly absorbed hard x - ray emitters .
they discussed many astrophysical processes including thermonuclear shell burning , wind collision and boundary layer in the accretion disk . in symbiotic stars
unique and useful spectroscopic diagnostics are provided from raman scattering by atomic hydrogen . in this scattering process
, a far uv photon blueward of ly@xmath10 is incident upon a hydrogen atom in the ground state .
subsequently the hydrogen atom de - excites into @xmath11 state re - emitting an optical raman - scattered photon .
the first identification of raman scattering in symbiotic stars was made by @xcite , who proposed that the broad emission features at 6825 and 7088 usually observed in many symbiotic stars are raman scattered o vi@xmath2 1032 and o vi@xmath2 1038 .
the operation of raman scattering requires the coexistence of a highly thick h i region and a strong far uv emission region , which is ideally met in symbiotic stars .
recently raman scattered o vi 1032 and 1038 features were found in the b[e ] star lha 115-s 18 ( @xcite ) .
raman spectroscopy is very important in probing the mass loss and mass transfer processes in symbiotic stars ( e.g. @xcite , @xcite ) . adopting an emission region that is in a keplerian motion around the hot component
, @xcite successfully fitted the double peaked profiles of the raman scattered o vi@xmath21032 of the symbiotic stars hm sge and v1016 cygni .
the representative keplerian velocity scale @xmath12 of the emission regions in these two symbiotic stars implies that the major o vi emission region is located roughly within one astronomical unit from the hot component .
it should be also noted that a double peak profile in raman scattered o vi can be produced from purely geometric effects , where the scattering region coincides with the slowly expanding stellar wind around the giant ( e.g. @xcite , @xcite )
. the mass loss rate of the cool component of v1016 cyg has been deduced from a photoionization model calculation combined with the measurement of the center shift of raman scattered he ii by @xcite .
additional raman scattered features were found and proposed by @xcite , who discussed the plausibility of raman scattering for various far uv emission lines including he ii and c iii . in particular , he reported the detection of the raman scattered he ii features blueward of h@xmath13 and h@xmath14 in the symbiotic star rr telescopii .
these features are also found in the symbiotic stars hm sge and v1016 cyg ( e.g. @xcite , @xcite , @xcite ) . in a be - like isoelectronic sequence , ne vii is a very highly ionized species with the ionization potential of 207.271 ev ( e.g. @xcite ) .
because of this high ionization potential , ne vii lines are reported only in extremely hot stellar objects and active galactic nuclei .
@xcite found ne vii@xmath2973 absorption lines in the spectra of pg 1159 stars obtained with the _ far ultraviolet spectroscopic explorer ( fuse)_. @xcite proposed the existence of ne vii@xmath2973 in the fuse spectrum of the symbiotic star ag draconis . in the
seyfert 1 galaxy ngc 5548 @xcite identified ne vii / ne viii blend at 88 in the extreme uv spectrum .
v1016 cyg is a d type symbiotic star with the cool component being a mira type variable .
it is also a symbiotic nova that underwent a nova - like outburst in 1964 ( @xcite ) . according to the investigation of @xcite , d type symbiotic novae including hm sge , rr tel and v1016 cyg form the subgroup of symbiotic stars with the hot components characterized by the highest temperature and luminosity .
v1016 cyg being of higher temperature than ag dra , this leads to an interesting possibility that v1016 cyg may show the emission line ne vii@xmath2973 . in this paper , we present the high resolution spectra of v1016 cyg , in which a broad emission feature at 4881 is found redward of h@xmath13 .
we propose this broad feature is formed through raman scattering of the far uv emission line ne vii@xmath2973 in the thick h i region .
obtained with boes on 2003 december 16 .
the lower panel is a blow - up version of the upper panel by a factor 24 .
the vertical axis shows the ccd counts . the broad feature around 4882 marked by the solid vertical arrow shown in the bottom panel
is proposed to be raman scattered ne vii@xmath2973 by atomic hydrogen . ]
[ boes_fig1 ] shows a high resolution spectrum of v1016 cyg obtained with the bohyunsan optical echelle spectrograph ( boes ) installed on the 1.8 m optical telescope at mt .
bohyun on the night of 2003 december 16 . in fig
. [ boes_fig2 ] , we show another boes spectrum of v1016 cyg obtained on 2005 november 7 . for the spectrum of 2003 , the 200 @xmath15 m optical fiber was used to yield the spectroscopic resolution of @xmath16 . on the other hand ,
the 2005 spectrum was obtained using the 300 @xmath15 m optical fiber achieving spectroscopic resolution of @xmath17 .
the exposure time was 3000 seconds and 7200 seconds for the spectrum of 2003 and that of 2005 , respectively . in fig .
[ boes_fig1 ] , the lower panel is a blow - up version of the upper panel by a factor 24 . in a similar way , in fig .
[ boes_fig2 ] the scale of the lower panel is 50 times smaller than that of the upper panel . in the 2005 spectrum ,
the emission line h@xmath13 is saturated in order to allow a close investigation of the weak features around h@xmath13 .
we also obtained another spectrum with a short exposure time of 300 seconds , which is shown in fig .
[ gas_fit1 ] . as in the analysis of @xcite we find raman scattered he ii@xmath2972 at 4850 in the two long exposure spectra
obtained in 2003 and 2005 , which is very broad and formed in the blue wing part of h@xmath13 .
because of energy conservation , the wavelength @xmath18 of the raman scattered radiation in vacuum is related to that of the far uv incident radiation @xmath19 by the equation @xmath20 where @xmath21 is the line center wavelength of hydrogen ly@xmath10 . with the refractive index of air @xmath3
, the observed wavelength will be reduced to @xmath22 in particular , the incident wavelength @xmath23 for he ii@xmath2972 gives rise to the raman scattered he ii at @xmath24 ( @xcite ) , which yields the observed wavelength of @xmath25 . in fig [ boes_fig1 ] , the broad raman scattered he ii feature has the observed line center at @xmath26 , which is blueshifted from @xmath27 by @xmath28 from the expected center wavelength . on the other hand , in fig .
[ boes_fig2 ] , we find that @xmath29 . this deviation can be explained that the he ii emission region is receding from the neutral h i region
. more detailed investigation on the kinematics of the emission region is described in the subsequent section .
however , line center shift can also occur due to the varying raman conversion rate dependent on the wavelength ( @xcite ) , which necessitates monte carlo simulations for more accurate determination of the kinematics of the he ii emission region relative to the h i region .
( [ raman_length ] ) also leads to the following relation @xmath30 which results in raman scattered features with a significantly broadened profile .
in particular , the raman scattered features around h@xmath13 are broadened by a factor @xmath31 , because @xmath32 and @xmath33 . using a gaussian function given by @xmath34 , \label{gaussian}\ ] ] we perform a line profile fitting analysis to determine the observed centers of the emission lines he ii@xmath24859 and h@xmath13 in the 2005 spectrum of v1016 cyg .
because of the saturation of h@xmath13 , we use another spectrum obtained on the same night with an exposure time of 300 seconds . in fig .
[ gas_fit1 ] , we show the result , which is also summarized in table 1 .
in a similar way , the raman scattered he ii@xmath2972 and the broad feature at 4883 are fitted for the spectrum with exposure time of 7200 seconds , of which the result is shown in fig .
[ gas_fit2 ] . for this , the broad h@xmath13 wings are fitted using a profile @xmath35 , where @xmath36 is the wavelength of h@xmath13 . in the case of the 2003 spectrum ,
the profile fitting was done by @xcite except the broad feature at 4881 . in this work ,
the observed line center of this feature in the 2003 spectrum is determined to be 4880.80 .
the result of our single gaussian fit to the 2003 spectrum is also summarized in table 1 .
4859 and h@xmath13 using a single gaussian function for the spectrum obtained on the night of 2005 november 7 with the exposure time of 300 seconds .
gaussian fitting functions are shown by dotted lines and the boes spectrum is represented by a solid line .
the boes data are fitted by combining the two gaussian functions , which is represented by a dot - dashed line .
the observed line center of he ii@xmath24859 is 4858.57 and that of h@xmath13 is 4860.54 . ]
972 and a broad feature at 4883 using a single gaussian function for the spectrum obtained on the night of 2005 november 7 with the exposure time of 7200 seconds . for the h@xmath13 wing the profile function proportional to @xmath35 is used for fitting , which is shown by a thick dotted line .
the boes spectrum is shown by a solid line and the two single gaussian functions are represented by a dotted line .
the dot - dashed line shows the combined fit of the two gaussians and h@xmath13 wing .
the observed line center of raman scattered he ii@xmath2972 is 4852.95 and that of raman scattered ne vii@xmath2973 is 4882.82 . ] both in figs .
[ boes_fig1 ] and [ boes_fig2 ] , the broad emission feature marked by a solid vertical arrow is noticeable .
this feature in the 2003 spectrum is very weak with the observed center at 4880.80 but it is significantly clear at 4882.82 in the spectrum of 2005 with much longer exposure time .
we propose that this broad feature is formed through raman scattering of ne vii@xmath2973 by atomic hydrogen .
ne vii@xmath0 is formed through recombination of ne viii accompanied by a radiative transition from @xmath37 to @xmath38 ( e.g. @xcite ) .
this line was also found in the high resolution coronal spectrum of the sun obtained with the solar ultraviolet measurement of emitted radiation ( sumer ) on board the _ solar and heliospheric observatory , soho _
( @xcite ) .
@xcite proposed that ne vii@xmath2973 is responsible for the strong p cygni features observed at around 975 for a number of hot evolved stars including a78 , ngc 2371 and k1 - 16 .
the atomic data for be - isoelectronic atoms can be found in the work of @xcite , who compared the theoretical values with the experimental data .
he recommended the values of 214954 @xmath39 and 317954 @xmath39 for the lower and the higher energy levels respectively , from which the center wavelength of ne vii@xmath2973 is obtained to be 973.350 .
however , according to @xcite , the energy values of the lower and the higher levels are 214951.6 @xmath39 and 317694.6 @xmath39 respectively , resulting in the central wavelength of 973.302 .
this result is based on the previous work of @xcite . in this work
we adopt this value as the line center wavelength @xmath40 of ne vii@xmath2973 . a direct substitution of the ne vii line center wavelength into eq .
( [ raman_length ] ) yields the vacuum line center wavelength of the raman scattered ne vii feature , which is given by @xmath41 .
considering the refractive index of air , this vacuum wavelength is reduced to @xmath42 from which we propose to call this broad feature raman scattered ne vii@xmath2973. in section 3 , we perform monte carlo simulations to fit the profile , from which we may deduce useful constraints on the strength and profile of the unobserved ne vii@xmath2973 .
this line center value @xmath43 is shorter than the observed center wavelength @xmath44 by a small amount of @xmath45 in the 2003 spectrum . in the case of the 2005 spectrum ,
the corresponding values are @xmath46 , and @xmath47 .
this feature is quite broad , which is consistent with the primary characteristic of a raman scattered feature .
a similar spectroscopic behavior is seen in the case of raman scattered he ii@xmath2972 . in the 2003 data , the observed line center of raman scattered he
ii@xmath2972 is almost coincident with that expected from the atomic line center .
in the 2005 spectrum it is redshifted by an amount exceeding 2 . because both he ii and ne vii are highly ionized species , it is quite plausible that their emission regions almost coincide sharing a similar kinematics with respect to the neutral region .
because the profiles of raman scattered features are dominantly affected by a relative motion between the far uv emission source and the h i region and little influenced by the observer s sightline , the difference in the spectra of 2003 and 2005 implies that there was a significant change in the far uv emission region and/or in the h i region of v1016 cyg . [ cols="^,^,^,^,^ " , ] 972 and ne vii@xmath2973 .
the adopted neutral hydrogen column density is @xmath48 . in the upper panel ,
we show the 2003 data by a solid and the monte carlo result by a dotted line . in the lower panel ,
the boes data is deleted in order to show clearly the monte carlo result .
for the 2003 data , both the he ii and ne vii emission regions are assumed to coincide and be at rest with respect to the h i region . ] . in this simulation ,
both the he ii and ne vii emission regions are assumed to coincide but recede from the h i region with a velocity @xmath49 . ]
the operation of raman scattering by atomic hydrogen is a very unique astrophysical process that distinguishes symbiotic stars from other celestial bodies . thus far
the raman scattered line features found in symbiotic stars are limited to only he ii and o vi . in this work
we propose that the broad feature around 4881 in the spectrum of v1016 cyg is formed through raman scattering of ne vii@xmath2973 .
we have provided the atomic physical quantities regarding raman scattering of ne vii@xmath2973 . by performing monte carlo simulations , we obtain the random velocity scale of ne vii and infer the equivalent width of the emission line ne vii@xmath2973 . assuming that both ne vii@xmath2973 and he ii@xmath2972 emission regions coincide , we find that both he ii and ne vii emission components are receding from the neutral region with the same speed @xmath50 in the 2005 boes data . however , the 2003 boes data indicate that both ne vii and he ii emission regions are consistent with their being at rest with respect to the h i region at that time .
we are grateful to an anonymous referee for very useful comments that improved the presentation of the paper .
we also thank the staff of the bohyunsan optical observatory for their help in securing the high resolution spectra of v1016 cyg .
this research was supported by the basic science research program through the national research foundation ( nrf ) funded by the ministry of education , science and technology ( 2011 - 0027069 ) .
angeloni , r. , contini , m. , ciroi , s. , rafanelli , p. , mnras , 402 , 2075 belczyski , k. , mikoajewska , j. , munari , u. , ivison , r. j. , friedjung , m. , 2000 , a&as , 146 , 407 bethe , h. a. & salpeter , e. e. , 1957 , quantum mechanics of one- and two - electron atoms , new york , academic press birriel j. , 2004 , apj , 612 , 1136 edln , b. , 1983 , physica scripta , 28 , 51 feldman u. , behring w. e. , curdt w. , schle u. , wilhelm k. , lemaire p. & moran k. t. m. , 1997 , apj , 113 , 195 harries , t. j. & howarth , i. d. , 1997 , a&as , 121 , 15 herald j. e. , bianchi l. & hillier d. j. , 2005 , apj , 627 , 424 iben , i. , jr . ,
tutukov , a. v. , apjs , 105 , 145 jung y .- c . &
lee h .- w . , 2004 , mnras , 355 , 221 kaastra , j. s. , roos , n. , & mewe , r. 1995 , a&a , 300 , 25 , kenyon , s. j. , 1986 , the symbiotic stars , cambridge university press kramida , a. & buchet - poulizac , m .- c . , 2006 , european physical journal d , 38 , 265 kramida , a. , ralchenko , yu . ,
reader , j. , and nist asd team , 2013 , _ nist atomic spectra database ( ver . 5.1 ) _ lee h .- w . , 2000 , apj , 541 , l25 lee h .- w .
, 2012 , apj , 750 , 127 lee h .- w .
, jung y .- c .
, song i .-
& ahn s. , 2006 , apj , 636 , 1045 lee h .- w . & kang s. , 2007 , apj , 669 , 1156 lee h .- w . &
, 1999 , apj , 515 , l89 luna , g. j. m. , sokoloski , k. , mukai , k. & nelson , t. , 2013 , a&a , 559 , a6 mastrodemos , n. , morris , m. , 1998 , apj , 497 , 303 mikolajewska , m. , baltic astronomy , 21 , 5 mccusky , s. , 1965 , iau circ .
1916 mrset , u. nussbaumer , h. , schmid , h. m. & vogel , m. , 1991 , a&a , 248 , 458 nussbaumer , h. , 2003 , aspc , 303 , 557 schmid , h. m. , 1989 , a&a , 211 , l31 schmid , h. m. , 1992 , a&a , 254 , 224 skopal , a. , 2006 , a&a , 457 , 1003 sokoloski , j. l. , bildsten , l. , & ho , w. c. g. , 2001 , mnras , 326 , 553 torres a. f. , kraus m. , cidale l. s. , barb r. , borges fernandes m. & brandi e. , 2012 , mnras , 427 , l80 van groningen , e. , 1993 , mnras , 264 , 975 warner , b. , 1995 , cataclysmic variable stars , cambridge university press , werner , k. , rauch , t. , reiff , e. , kruk , j. w. & napiwotzki , r. , 2004 , a&a , 427 , 685 young , p. r. , dupree a. k. , espey b. r. , kenyon , s. j. , ake , t. b. , 2005 , aj , 618 , 891 | we present the high resolution spectra of the symbiotic star v1016 cygni obtained with the bohyunsan optical echelle spectrograph in 2003 and 2005 , from which we find a broad emission feature at 4881 .
we propose that this broad feature is formed from raman scattering of ne vii@xmath0 by atomic hydrogen .
thus far , the detection of raman scattered lines by atomic hydrogen is limited to o vi@xmath11032 , 1038 and he ii@xmath1940 , 972 and 1025 . with
the adoption of the center wavelength 973.302 of ne vii@xmath2973 and consideration of the air refractive index of @xmath3 , the atomic line center of the raman scattered ne vii feature is determined to be 4880.53 .
the total cross section at the line center of ne vii@xmath0 is computed to be @xmath4 with the branching ratio of 0.17 .
we perform monte carlo simulations to fit the raman scattered ne vii@xmath0 . assuming that the ne vii and he ii emission regions share the same kinematics with respect to the neutral scattering region , we find that the raman scattered he ii@xmath2972 at 4850 and ne vii@xmath2973 at 4881 are excellently fitted .
we also propose that the he ii and ne vii emission regions were stationary with respect to the h i region in 2003 but that they were receding from it with a velocity @xmath5 in 2005 .
[ firstpage ] binaries : symbiotic atomic processes line : identification scattering stars ( individual v1016 cyg ) |
superconductivity with a remarkably high transition temperature @xmath2 k was recently discovered in mgb@xmath3 @xcite . the subsequent investigation of the boron isotope effect by budko _ et al . _
@xcite revealed a partial isotope exponent @xmath4 ( corresponding to @xmath5k ) which appears to be consistent with a phonon - mediated bcs superconducting ( sc ) mechanism .
also other experimental reports as e.g.on specific heat @xcite argued that their data can be accounted for by a conventional , s - wave type bcs - model .
investigations of the sc gap of @xmath0 by means of raman scattering are also consistent with an isotropic s - wave gap with a moderate coupling @xmath6 @xcite . however , a theoretical analysis of the temperature dependence of the upper critical field @xmath7 in terms of eliashberg type models by shulga _ et al . _
@xcite demonstrates that the shape and the magnitude of the upper critical field of mgb@xmath3 can definitely not be accounted for by an isotropic single band model , but may successfully be described within a multi - band eliashberg model with various options . a careful calorimetric investigation of the sc parameters of mgb@xmath3 by wang _ et al .
@xcite gave even more direct evidence against the arguments for simple isotropic bcs type superconductivity : `` the nearly quadratic dependence of @xmath8 versus @xmath9 at @xmath10 , its non - linear field dependence , and the discrepancy between the electron - phonon coupling constant @xmath11 as determined by the renormalization of the electron density of states ( @xmath12 ) and by mcmillan s equation for isotropic superconductors ( @xmath13 ) , are inconsistent with a single isotropic gap '' .
direct hints for a non bcs temperature dependence of the gap energy @xmath14 were also obtained by tunneling experiments on mgb@xmath3/ag and mgb@xmath3/in junctions @xcite .
thermal conductivity @xmath15 is one of those transport coefficients which exhibits non - zero values in both the normal and the sc state .
the temperature dependence of @xmath15 allows to distinguish between the most important interactions present in a superconductor .
in particular , the interaction of electrons with phonons are recorded in the magnitude of @xmath16 .
moreover , scattering of these particles by static imperfections like impurities , defects or grain boundaries are reflected .
the aim of the present work is to derive the temperature dependent thermal conductivity of @xmath0 and to analyse the data with respect to the electronic - and the lattice thermal conductivity both in the normal and the sc state .
moreover , we present resistivity and specific heat measurements in order to characterise the investigated sample .
a @xmath0 sample of about 1.3 g was synthesized by direct reaction of the elements .
the starting materials were elemental magnesium ( rod 99.9 mass % nominal purity ) and boron ( 99.5 % powder , crystalline , @xmath17 60 mesh , 99.5 mass % ) .
the elements in a stoichiometric ratio were enclosed in a cylindric tantalum crucible sealed by arc welding under argon atmosphere .
the tantalum crucible was then sealed in an iron cylinder and heated for one hour at @xmath18 c and two hours at @xmath19 c in a furnace .
the sample characterized by x - ray diffraction show pure @xmath0 phase ; only one very weak peak due to extra phase has been found .
the thermal conductivity measurement was performed in a flow cryostat on a cuboid - shaped sample ( length : about 1 cm , cross - section : about 2.5 mm@xmath20 ) , which was kept cold by anchoring one end of the sample onto a thick copper panel mounted on the heat exchanger of the cryostat .
the temperature difference along the sample , established by electrical heating , was determined by means of a differential thermocouple ( au + 0.07 % fe / chromel ) .
the measurement was performed under high vacuum and 3 shields mounted around the sample reduced the heat losses due to radiation at finite temperatures .
the innermost of these shields is kept on the temperature of the sample via an extra heater maintained by a second temperature controller .
resistivity data were taken from bar shaped samples applying a standard 4-probe d.c .
technique at temperatures down to 0.5 k and in magnetic fields up to 12 t. specific heat measurements were carried out on a sample of about 1 g in the temperature range 5k50k using a quasi adiabatic step heating technique .
in order to give a direct proof of the sc bulk properties of our mgb@xmath3 sample prior to the transport measurements we checked the specific heat .
these specific heat measurements performed in zero - field , 1 and 9 t showed reasonable agreement with results previously reported @xcite .
the thermodynamic mean transition temperature @xmath21 of our sample is 37.5k .
as already noted in the introduction there have been distinctly different and partly controversial conclusions suggested in refs .
@xcite , although their raw data are in fair agreement with each other .
therefore , we show in figure [ fig0 ] the temperature dependence of the electronic specific heat , @xmath22 versus @xmath23 obtained by subtracting the lattice heat capacity deduced from the 9 t specific heat data . of course , 9 t are insufficient to obtain a complete suppression of superconductivity in mgb@xmath3 , but there is already a large reduction of the order parameter combined with a dramatic broadening of the transition ( see ref .
@xcite for comparison of 10 , 14 and 15 t data ) .
thus , we obtained a sommerfeld coefficient of the normal - state electronic specific heat @xmath24(2 ) mj / molk@xmath20 by extrapolating the 9 t data in @xmath25 versus @xmath26 from 30 100 k@xmath20 to zero temperature which is already close to @xmath27 mj / molk@xmath20 obtained from the 14 and 16 t measurement by wang _
et al . _
the important point to emphasize in figure [ fig0 ] is the non bcs - like temperature dependence of the sc - state electronic specific heat .
in fact , we observed a similar deviation from a simple bcs temperature dependence ( solid line , figure [ fig0 ] ) as previously reported by wang _
et al . _
@xcite ( @xmath28 $ ] for @xmath29 ) .
this discrepancy is supposed to be indicative for the opening of an additional gap below about 10k .
the solid line in figure 1 indicates that a fraction of electrons corresponding to a normal state @xmath30 mj / molk@xmath20 is tentatively accounted for by the bcs - fit with @xmath31 while a second fraction corresponding to @xmath32 mj / molk@xmath20 contributes to the smaller gap opening at temperatures well below @xmath33 to further screen the quality of the sample , temperature and magnetic field dependent resistivity measurements @xmath34 were performed from 0.5 k up to room temperature . shown in figure [ fig1 ] ( right axis )
is @xmath35 at zero field .
the transition into the sc state occurs at @xmath36 k , which is in fine agreement with already published data .
the rrr ratio of this polycrystalline sample is about 6 .
the resistivity behaviour in the normal state region matches a dependence according to @xmath37 with the residual resistivity @xmath38 and the coefficient @xmath39 .
a @xmath26 behaviour of @xmath35 was recently reported @xcite and it seems to reflect interactions between charge carriers .
a study of the hall coefficient implies that electrical transport is dominated by holes @xcite .
the value of the coefficient @xmath40 , however , is significantly smaller than that known e.g. , for highly correlated electron systems , but seems to reflect the modest density of states at fermi energy @xcite .
it should be mentioned that other power laws with an exponent close to 3 were reported for sintered material of @xmath0 @xcite .
measurements of the resistivity down to 0.5 k and in fields up to 12 t indicate that @xmath41 is well above that limit .
interestingly , the transition region becomes much broader with increasing fields , but different values of imprinted currents ( from 5 to 40 ma ) do not change the width of the transition .
the temperature dependent thermal conductivity @xmath15 of @xmath0 is shown in figure [ fig1 ] .
the overall behaviour of @xmath42 is typical of an intermetallic compound where scattering on static imperfections prohibits a pronounced maximum to occur , as it is the case in pure and simple metals .
moreover , the absolute magnitude appears to be of that order , as usually found for intermetallics .
anomalous behaviour of @xmath16 in the proximity of @xmath1 is not observed and a local maximum or a pronounced shoulder below @xmath1 do not occur in the investigated sample .
generally , the total thermal conductivity of metals consists of a sum of an electronic contribution @xmath43 and a lattice contribution @xmath44 : @xmath45 in order to separate both contributions from the total measured effect , the wiedemann - franz law is applied , assumed to be valid , at least , in simple metals .
this model relates the electrical resistivity @xmath46 with the electronic contribution to the thermal conductivity @xmath43 and can be expressed as @xmath47 where @xmath48 @xmath49 is the lorenz number .
using equation [ eq1 ] and taking into account the appropriate values of the electrical resistivity in the normal state region of @xmath0 ( compare figure [ fig1 ] , right axis ) allows to split @xmath15 into @xmath50 ( dashed line , figure [ fig1 ] ) and @xmath44 ( dashed - dotted line , figure [ fig1 ] ) .
this type of analysis indicates that both contributions are almost equal in the entire temperature range of the normal state region of @xmath0 . according to matthiessen s rule
both @xmath43 and @xmath44 are limited owing to various scattering processes , which can be expressed in terms of a thermal resistivity @xmath51 . in the case of non - magnetic materials , the following temperature dependence of the electronic contribution to the total measured quantity
is assumed to be valid @xcite : @xmath52 where the subscripts ( e,0 ) and ( e , ph ) refer to interactions of the conduction electrons with static imperfections and thermally excited phonons , respectively ; @xmath53 and @xmath54 are material constants .
equation [ eq2 ] allows to detemine @xmath55 and @xmath56 . shown in figure [ fig2 ]
is the electronic thermal resistivity @xmath57 of @xmath0 displayed in the normal state region up to about 80 k. the solid line is a least squares fit of the data according to equation [ eq2 ] and the dashed and the dashed - dotted lines represent @xmath55 and @xmath56 , respectively .
thus , the deduced parameters are @xmath58 and @xmath59 . obviously from figure [ fig2 ] ,
the scattering of electrons with static imperfections of the crystal becomes dominant as the temperature approaches @xmath1 .
the relative weight of @xmath43 and @xmath44 at the sc transition temperature @xmath1 , as defined from the wiedemann franz law , also serves to determine the temperature dependence of @xmath43 and subsequently of @xmath44 below @xmath1 .
since within the bcs theory , cooper pairs do not carry heat and entropy , the scattering terms of equation [ eq2 ] have to be modified in order to account for the decreasing number of unpaired electrons .
in the sc state , the thermal resistivity @xmath60 can be represented as @xmath61 with @xmath62 @xcite .
the functions @xmath63 and @xmath64 were calculated repeatedly and agree well with experimental findings @xcite . in the dirty limit of a superconductor , the first term of equation [ eq3 ]
dominates , i.e. , @xmath65 . in terms of the brt theory
@xcite @xmath66 is a universal function of @xmath67 , dependent on the value - and the temperature dependence of the sc gap @xmath68 . on the contrary ,
clean limit superconductors are dominated by the second term of equation [ eq3 ] , revealing @xmath69 .
geilikman et al .
@xcite have calcutated and tabulated @xmath64 which yields again a universal behaviour on @xmath67 . differently to scattering on imperfections , @xmath70 in the bcs limit increases initially with decreasing values of @xmath67 in spite of a rapid decrease of electronic excitations .
a maximum occurs at @xmath71 with @xmath72 .
various high temperature superconductors are found to exhibit a maximum in @xmath16 below @xmath1 and thus the origin of that feature is , at least partly , attributed to a significant scattering strength of electrons on thermally excited lattice vibrations ( compare e.g. @xcite ) .
figure [ fig2 ] evidences that slighty above @xmath1 , @xmath55 exceeds @xmath56 by more than one order of magnitude and contributes at this temperature about 95 % to @xmath73 .
this implicitely favours a description of the thermal conductivity of @xmath0 based on scattering of electrons on impurities .
nevertheless , for the present analysis of the data both terms of equation [ eq3 ] are considered in order to analyse the electronic contribution to the total thermal conductivity .
taking into account the functions @xmath63 and @xmath64 and the numerical values of @xmath53 and @xmath54 allows us to determine @xmath74 below @xmath1 ( compare figure [ fig3 ] , panel ( a ) ) .
@xmath74 decreases with decreasing temperature and is primarily determined by the brt function @xmath63 .
the difference @xmath75 represents the phonon - originated thermal conductivity @xmath76 in the sc state .
the latter appears to be larger in the sc state than @xmath74 . for temperatures @xmath77 , the lattice thermal conductivity of @xmath0 becomes dominant
this nicely agrees with theoretical considerations @xcite .
the lattice term @xmath76 is constrained by various scattering processes ; among them are interactions of the phonons with electrons , point defects , dislocations or sheetlike faults . to account for observed deviations of the sc gap of @xmath0 from the bcs theory , the experimental data from tunnel experiments on @xmath78 and @xmath79 @xcite have been used to modify the function @xmath63 and thus @xmath80 .
the smaller the value of @xmath81 with respect to the bcs theory , the less steep is the decrease of @xmath63 when the temperature is lowered . since @xmath82 of @xmath0 as obtained from that study ( @xmath83 , @xcite ) is well below the bcs value ( @xmath84 ) , @xmath74 - in this type of analysis - becomes larger in the sc state down to @xmath85 .
still , @xmath86 .
taking @xmath76 as derived from the bcs - like gap , however , provides a slightly smoother crossover from the sc - to the normal state region of @xmath0 .
panel ( b ) of figure [ fig3 ] shows the temperature dependent thermal resistivity @xmath87 .
here , the data derived from the bcs model are used .
obviously , scattering of electrons by phonons in the sc state contributes just a fraction to the thermal resistivity @xmath88 , and therefore @xmath80 is the most significant term below @xmath1 .
this , of course , will not change if the actual dependence of @xmath14 is considered .
the measurement of the thermal conductivity of @xmath0 reveals no pronounced anomaly at @xmath89 and furthermore no local maximum occurs at temperatures well below @xmath1 .
such a behaviour is most likely caused by the dominance of electron ( hole ) scattering on static imperfections present in the investigated @xmath0 sample .
in fact , an analysis based on the brt model and the model of geilikman evidences that @xmath80 is the predominant term . beside the classical bcs behaviour of the sc gap
, an attempt was made to incorporate the actual gap behaviour reported for @xmath0 @xcite .
the subsequent analysis shows that the electronic contribution to the thermal conductivity becomes larger than in the former case .
the application of the wiedemann franz law to the experimental data indicates that both the electronic - and the lattice contribution to the total thermal conductivity are almost of the same size over a large temperature range .
this work is supported by the austrian fwf p12899
. 99 j.nagamatsu , n. nakagawa , t. muranaka , y. zenitani , and j. akimitsu , nature * 410 * , 63 ( 2001 ) .
budko , g. lapertot , c. petrovic , c.e .
cunningham , n. anderson , and p.c .
canfield , phys .
lett . * 86 * , 1877 ( 2001 ) .
kremer , b.j .
gibson , and k. ahn , cond - mat/0102432 ( 2001 ) .
wlti , e. felder , c. degen , g. wigger , r. monnier , b. delley , and h.r .
ott , cond - mat/0102522 ( 2001 ) .
chen , m.j .
konstaninovic , j.c .
irvin , d.d .
lawrie and j.p .
franck , cond - mat/0104005 ( 2001 ) .
shulga , s .- l .
drechsler , h. eschrig , h. rosner , and w. pickett , cond - mat/0103154 ( 2001 ) .
y. wang , t. plackowski , and a. junod , cond - mat/0103181 ( 2001 ) .
a. plecenik , s. benacka , and p. kus , cond - mat/0104038 ( 2001 ) .
jung , m.s .
park , w.n .
kang , m.s .
kim , s.y .
lee , and s.i .
lee , cond - mat/0102215 ( 2001 ) .
w. n. kang , c. u. jung , k.h.p .
kim , m.s .
park , s.y .
lee , h.j .
kim , e.m .
choi , k.h .
kim , k.m.s .
kim , and s.i .
lee , cond - mat/0102313 ( 2001 ) .
j. kortus , i.i .
mazin , k.d .
belashenko , v.p .
antopov , and l.l .
boyer , cond - mat/0101446 ( 2001 ) .
finnemore , j.e .
ostenson , s.l .
budko , g. lapertot , and p.c .
canfield , phys .
lett . * 86 * , 2420 ( 2001 ) .
canfield , d.k .
finnemore , s.l .
budko , j.e .
ostenson , g. lapertot , c.e .
cunningham , and c. petrovic , phys .
lett . * 86 * , 2423 ( 2001 ) .
klemens , in _ thermal conductivity _ , vol .
1 , ed . r.p .
tye , ( academic press , london and new york 1969 ) .
geilikman , m.i .
dushenat , and v.r .
chechetkin , sov .
jetp * 46 * , 1213 ( 1977 ) .
j. bardeen , g. rickayzen , and l. tewordt , phys . rev . * 113 * , 982 ( 1959 ) .
l. tewordt , phys . rev . * 129 * , 657 ( 1963 ) .
s. castellazzi , m.r .
cimberle , c. ferdeghini , e. giannini , g. grasso , d. marre , m. putti , and a.s .
siri , physica c * 273 * , 314 ( 1997 ) .
@xmath90 versus @xmath9 obtained by subtracting the lattice contribution , @xmath91 where @xmath92 mj / molk@xmath93 the full line indicates a bcs temperature dependence of @xmath94 as explained in the text.,width=453,height=340 ] ( left axis ) , and electrical resisitivity @xmath35 ( right axis ) of @xmath0 . the dashed and the dashed - dotted lines are the electronic - and the lattice contributions @xmath43 and @xmath44 , respectively.,width=377,height=245 ] of @xmath0 in the normal state region .
the dashed and the dashed - dotted lines are the contributions due to electron - imperfection - and electron - phonon scattering @xmath55 and @xmath56 , respectively.,width=377,height=245 ] of @xmath0 in the sc state .
the solid and the dashed - dotted lines represent electron- and lattice contributions @xmath74 and @xmath76 , derived from the bcs - and the modified model as discussed in the text , respectively .
( b ) : temperature dependent electronic thermal resistivity , @xmath95 of @xmath0 in the sc state derived from the bcs model .
the dashed and the dashed - dotted lines are the contributions due to electron - imperfection - and electron - phonon scattering @xmath80 and @xmath96 , respectively.,width=529,height=377 ] | thermal conductivity of superconducting @xmath0 was studied in both the superconducting and the normal state region .
the latter is almost equally determined by the electronic - and the lattice contribution to the total thermal conductivity . in the superconducting state , however , the lattice contribution is larger .
the electronic thermal conductivity below @xmath1 was derived from the experimental data considering the bardeen - rickayzen - tewordt theory together with the model of geilikman .
the analysis shows that electron scattering on static imperfections dominates . * key words
* : @xmath0 superconductivity , thermal conductivity , specific heat |
the structure of complex systems across various disciplines can be abstracted and conceptualized as complex networks of nodes and links to which many quantitative methods can be applied so as to extract any characteristics embedded in the system@xcite .
there exist many types of networks and characterizing their topology is very important for a wide range of static and dynamic properties .
recently , c. song _ _ et al.__@xcite applied a least box number covering algorithm to demonstrate the existence of self - similarity in many real networks .
they also studied and compared several possible least box number covering algorithms , by applying them to a number of model and real - world networks@xcite .
they found that the least box number covering optimization is equivalent to the well - known vertex coloring algorithm which is a np - hard problem .
it implied that any least box number covering algorithms are heuristic algorithms .
can we avoid the np - hard problem and simplify the network fractal dimension calculation ? in this paper we modify the random sequential box - covering algorithm@xcite and presented a random ball coverage algorithm .
the relative error smallest upper bound ( supremum ) is given theoretically .
the simulation experiments shows that no matter how large the network diameter is , this upper bound tends to @xmath1 .
so we develop another algorithm for calculating dimension which employ only two random ball coverages .
we also yield the smallest relative error upper bound of this algorithm , which tend to @xmath0 when the network diameter is large enough . in this point of view , for a proper acceptable error range , the random ball covering algorithm is equivalent to the least box number covering algorithm in statistic sense when a network is large enough .
there is no need to focus on the least box number covering optimization problem when we want to calculate dimension of a large diameter network .
the least box number coverage@xcite and random ball coverage were defined as following . for a given network @xmath2 ,
a box with diameter @xmath3 is a set of nodes where all distances @xmath4 between any two nodes @xmath5 and @xmath6 in the box are smaller than @xmath3 .
the least box number coverage is the box coverage with the minimum number of boxes required to cover the entire network @xmath2 . in order to correspond to the fractal network
definition@xcite we use open ball to cover the network .
so our random ball coverage is little difference with the random sequential box - covering algorithm@xcite .
a ball with radius @xmath3 and center node @xmath7 is the set of nodes which satisfy the shortest path length from the center @xmath7 to each of them is smaller than @xmath3 . the random ball coverage with radius @xmath3 as : at each step
, we randomly choose a node which has not been covered as a center , and cover all the nodes within the distance @xmath3 to the center .
the process is repeated until all the nodes in the network were covered .
theorem @xmath8 : @xmath9 @xmath10 , where @xmath11 is the number of boxes in a lest box number coverage with diameter @xmath3 and @xmath12 denotes the number of balls in a random ball coverage with radius @xmath3 .
proof : @xmath13 can be regarded as the number of boxes in a random box number coverage with diameter @xmath14 .
@xmath15 suppose @xmath16 denote all the boxes in a lest box number coverage , where @xmath17 , and @xmath18 .
then we have @xmath19 for all @xmath20 , where @xmath21 denotes empty set . according to the above definition of random ball coverage with radius @xmath3 , without loosing any generality
suppose the center of the first ball in a random ball coverage process is @xmath22 , then @xmath23 is covered by the first ball and the second random ball s center must lies out of @xmath23 . without loosing any generality
we also can assume @xmath24 is the center of the second ball , then @xmath25 is covered by the second ball .
in this way , we can get that : @xmath10 .
@xmath26 @xmath27 @xmath10 theorem @xmath28 : suppose @xmath2 is a fractal network , the available box diameter range is @xmath29 and we employ linear lest squares regression to get the dimension .
then the smallest upper bound of the relative error of dimension calculated by random ball coverage is @xmath30}{(r - m+1)\sum_{i = m}^{r}{(\log{i})^{2}}-(\log\frac{r!}{m!})^2}$ ] where , @xmath31 and satisfy : @xmath32 proof : @xmath33 network @xmath2 is fractal in range @xmath29 .
@xmath34 there must exist a proper @xmath35 for any @xmath36 such that @xmath37 @xmath38 then @xmath39\ ] ] where , @xmath40 is the dimension of the network .
suppose @xmath41 @xmath42 @xmath43 then @xmath44 and @xmath45}{(r - m+1)\sum_{i = m}^{r}{(\log{i})^{2}}-(\log\frac{r!}{m!})^2}$ ] @xmath34 the relative error is @xmath46}{(r - m+1)\sum_{i = m}^{r}{(\log{i})^{2}}-(\log\frac{r!}{m!})^2}$ ] employing linear programming , the upper bound of the relative error is:@xmath30}{(r - m+1)\sum_{i = m}^{r}{(\log{i})^{2}}-(\log\frac{r!}{m!})^2}$ ] where , @xmath31 and satisfy : @xmath32 fig.[relative_error_plot ] show the relationship among @xmath47 .
it is not easy to get any conclusion directly about the upper bound from the above expression .
so we have done some numerical calculations .
it seems that for a given @xmath48 when @xmath49 becomes large , it has a limit .
for instance , when @xmath48=1 or 2 , its limit is about @xmath1 .
it implies that no matter how large the network diameter is , the relative error never be lower than @xmath1 .
in fact , we could only use two points @xmath50 and @xmath51 to estimate dimension @xmath40 which is named twice - random ball coverage algorithm .
the corresponding upper bound of relative error will be tend to @xmath0 when the fractal network diameter tend to infinite .
theorem @xmath52 : suppose @xmath2 is a fractal network , the available box diameter range is @xmath29 .
we just use @xmath50 and @xmath51 to estimate network dimension
. then the estimated dimension value is @xmath53 and the smallest relative error upper bound is @xmath54 .
proof : obviously , @xmath55 and @xmath56 , then @xmath57 more details are shown in fig.[relative_error_plot ] . employing random ball coverage algorithm ( twice - random ball coverage algorithm ) , we get the fractal dimension of the world - wide web is @xmath58 , @xmath59diameter , @xmath60 , which is corresponding to the dimension obtained by c. song _
et al._(dimension is @xmath61)@xcite . from our empirical results ,
we find for many networks , @xmath62diameter and @xmath63 is more reasonable .
when @xmath63 we get www network dimension is @xmath64 .
sometimes , the available box diameter @xmath49 is not sufficient enough , we can calculate @xmath12 many time and get the dimension .
we also test this method in the @xmath65 cellular networks@xcite . for each network @xmath66=diameter , @xmath67 and each @xmath12 ,
we perform random coverage @xmath68 times .
then we get the average dimension of the whole cellular networks is @xmath69 which is perfect corresponding to the dimension obtained by c. song _
et al._(dimension is @xmath70)@xcite and w. zhou _ et al . _
( dimension is @xmath71)@xcite .
because we calculate each @xmath72 times , for any one of the @xmath65 cellular networks , we can get @xmath68 different dimensions . for each cellular network
we can get an average variance .
the maximum average variance of @xmath65 cellular network dimensions is @xmath73 , the average variance @xmath74 .
if we use the network dimension which is obtained by @xmath68 times calculation to substitute its real fractal dimension @xmath40 in our above discussion , we get the maximum relative error of @xmath68 time calculations of the @xmath65 cellular networks is @xmath75 .
the average maximum relative error is @xmath76 and the average relative error is @xmath77 .
the relative errors of empirical results are far less than the theoretical upper bounds respectively .
the interesting thing is that , in our theoretical discussion , the upper bound of twice - random ball coverage is less than the upper bound of random ball coverage algorithm .
but the empirical results always show the random ball coverage algorithm is better than twice - random ball coverage algorithm .
so , we think the random ball coverage algorithm is better than twice - random ball coverage algorithm in practice .
moreover , our theorems also can be used to estimate a network s diameter .
denotes the diameter of a network , @xmath78 denote the relative errors of random ball coverage algorithm and twice - random ball coverage algorithm.,width=302 ]
in this paper , we strictly present the upper bound of the relative error of random ball coverage method in fractal network dimension calculation .
and we also yield a simple relative error upper bound @xmath54 of twice - random ball coverage method . for many real - world networks ,
when the network diameter is sufficient enough this kind of relative error upper bound will tend to @xmath0 .
therefore , if the network is sufficient enough , twice - random ball coverage is equivalent to the leat box number coverage in fractal dimension calculation and calculating fractal network dimension is not a np - hard problem . for the networks which is
not sufficient enough , we can calculate random ball number many times and get the dimension , which is also very effective and accuracy .
the above discussions can lead another problem naturally .
we also can define random full box coverage .
a full box with diameter @xmath3 is a set of nodes , such that any other nodes out of the box is added to the box will make the box diameter larger or equal to @xmath3 .
the random full box coverage algorithm with diameter @xmath3 can be defined as@xcite : at each step we randomly choose a uncovered node @xmath79 as the first node of the box , and select the uncovered nodes to the box until the box become full .
we guess the random full box coverage algorithm is equivalent to the least box number covering algorithm in statistic sense . in the future
we will do some deep researches about this problem .
the authors want to thank chaoming song , qiang yuan for provide some useful information .
this work is partially supported by 985 projet and nsfc under the grant no.@xmath80 , no.@xmath81 and
no.@xmath82 . 99 r. albert , a .-
barabasi , rev .
phys . * 74 * , 47 ( 2002 ) .
m. e. j. newman , siam rev .
* 45 * , 167 - 256 ( 2003 ) .
s. boccaletti , v.latora , y. moreno , m. chavez , and d .- u .
hwang , physics report .
* 424 * , 175 - 308 ( 2006 ) .
song c , havlin s and makse h a , nature * 433 * 392 ( 2005 ) .
song c , havlin s and makse h a , nature physics * 2 * 275 ( 2006 ) .
c. song , l. k. gallos , s. havlin and h. a. makse , j .
stat . mech.p03006 ( 2007 ) .
j. s. kim , k .-
goh , b. kahng , and d. kim .
arxiv : cond - mat/0701504 ( 2007 ) . h. jeong , b. tombor , r. albert , z. n. oltvai and a .-
barabasi , nature * 407 * 651 - 654 , ( 2000 ) .
w. x. zhou , z. q. jiang , d.r sornette .
arxiv : cond - mat/0605676 ( 2006 ) . | least box number coverage problem for calculating dimension of fractal networks is a np - hard problem .
meanwhile , the time complexity of random ball coverage for calculating dimension is very low . in this paper
we strictly present the upper bound of relative error for random ball coverage algorithm .
we also propose twice - random ball coverage algorithm for calculating network dimension .
for many real - world fractal networks , when the network diameter is sufficient large , the relative error upper bound of this method will tend to @xmath0 . in this point of view , given a proper acceptable error range , the dimension calculation is not a np - hard problem , but p problem instead . key words : fractal network , dimension , random ball coverage pacs : 89.75.hc , 89.75.da , 05.45.df |
one of the most significant developments in the study of extragalactic globular cluster systems ( gcss ) was the discovery of bimodality in their color distributions ( see @xcite and references therein ) . today
, we generally refer to globular clusters ( gcs ) belonging to the blue peak of the color distribution as metal - poor gcs and to the red - peak members as the metal - rich sub - population .
it is generally considered that the presence of multiple modes implies multiple distinct gc formation epochs and/or mechanisms and ties those directly into formation scenarios that have to describe the _
parallel _ assembly histories of gcss _ and _ the diffuse stellar populations in their host galaxies . in massive early - type galaxies the current gcs assembly paradigms view the origin of the two color peaks from the perspective of either episodic star - cluster formation bursts triggered by gas - rich galaxy mergers ( e.g. * ? ? ?
* ) , temporarily interrupted cluster formation ( so - called _ in - situ _ formation , e.g. * ? ? ?
* ; * ? ? ?
* ) , and star - cluster accretion as a result of the hierarchical assembly of galaxies ( e.g. * ? ? ?
while the majority of gcss in early - type galaxies show clearly bimodal color distributions , the general picture is much more complex , ranging from purely blue to purely red color distributions ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
this complexity is exacerbated by the fact that color bimodality is a function of galaxy mass and morphology , as less massive and later - type galaxies tend to have single - mode blue ( i.e. metal - poor ) gc populations ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
furthermore , color bimodality is also a function of galactocentric distance and is mainly due to the more extended spatial distribution of the metal - poor sub - population relative to metal - rich clusters ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the aspect of gcs formation and assembly entered recently the domain of numerical simulations of galaxy formation due to the increasing spatial resolution of these computations .
for instance , @xcite model gc formation by identifying absorbing sink particles in their smoothed particle hydrodynamics ( sph ) high - resolution simulation of isolated gaseous disks and their mergers .
they find a bimodal globular - cluster metallicity distribution in their merger remnant under the assumption of a particular age - metallicity relation .
a key finding of their merger simulation is a more concentrated spatial distribution of metal - rich gcs with respect to the metal - poor sub - population in good agreement with observations . since their models of isolated galaxies produce a smooth age distribution ( implying a smooth metallicity and color distribution ) , @xcite conclude that mergers are required to produce a bimodal metallicity ( i.e. color ) distribution . in a more detailed adaptive - grid cosmological simulation
, @xcite followed the formation of a star - cluster system during the early evolution of a milky way - size disk galaxy to redshift @xmath1
. their model could reproduce the extended spatial distribution of metal - poor halo globular clusters as observed in m31 and the milky way . however , because their simulation does not follow the later evolution at @xmath2 it is unclear whether it would produce a metallicity bimodality and any significant age - metallicity relation .
an alternative , more statistical approach to modeling gcs assembly is to directly link the mode of gc formation to the star - formation rate in semi - analytic models .
@xcite were the first to explore this path by assuming that metal - poor gcs form in gaseous proto - galactic disks while metal - rich gcs are created during gaseous merger events .
their study showed that the observed globular - cluster color bimodality can only be reproduced by artificially stopping the formation of metal - poor gcs at redshifts @xmath3 . by construction , no spatial information on metal - rich and/or metal - poor gcs is provided in these models .
recently , ( * ? ? ?
* hereafter pm04 ) presented a spatially resolved chemical evolution model for the formation of spheroids , which successfully reproduces a large number of photo - chemical properties that could be inferred from either the optical or from the x - ray spectra of the light coming from ellipticals .
the model includes an initial gas infall and a subsequent galactic wind ; it takes into account detailed nucleosynthesis prescriptions of both type - ii and ia supernovae as well as low and intermediate - mass stars .
it has been extensively tested against the main photo - chemical properties of nearby ellipticals , including the observed increase of the @xmath0-enhancement in their stellar populations with galaxy mass ( e.g. * ? ? ?
* ; * ? ? ?
this is at variance with standard models based on the hierarchical merging paradigm , which do not reproduce this trend @xcite .
since the pm04 model provides full radial information on the composite nature of stellar populations that make up elliptical galaxies , the observation of different gc sub - populations is , therefore , a new sanity check for the validity of this model .
moreover , we recall that pm04 and , more recently , @xcite ( @xcite , hereafter pmc06 ) suggested that elliptical galaxies should form outside - in , namely the outermost regions form faster as well as develop an earlier galactic wind with respect to the central parts ( see also * ? ? ?
this mechanism implies that the stars in massive spheroids form a composite stellar population ( csp ) , whose chemical properties , in particular their metallicity distribution , changes with galactocentric distance .
starting with the assumption that gc sub - populations trace the components of csps , we will show how the observed multi - modality in gcss can be ascribed to the radial variation in the underlying stellar populations .
in particular , the observed gcss are a linear combination of gc sub - populations inhabiting a given projected galactocentric radius .
the paper is organized as follows : in section 2 we briefly describe the adopted theoretical model ; in section 3 we compare the predictions with observations and discuss the implications , while section 4 presents the final conclusions .
the chemical evolution code for elliptical galaxies adopted here is described in pm04 , where we refer the reader for more details . in this work
, we present the results for a galaxy with @xmath4 , taken from pm04 s model iib .
this model is characterized by a @xcite imf , @xcite yields for massive stars , @xcite yields for type - ia sne , and @xcite yields for low- and intermediate - mass stars .
an important feature of the pm04 model is its multi - zone nature , namely the model galaxy is divided into several non - interacting spherical shells of radius @xmath5 , which facilitate a detailed study of the radial variation of the photo - chemical properties of the gcs and its host galaxy . in each zone
_ i _ , the equations for the chemical evolution of 21 chemical elements are solved ( see pm04 , @xcite ) .
the model assumes that the galaxy assembles by merging of gaseous lumps on short timescales .
the chemical composition of the lumps is assumed to be primordial .
in fact , our model assumes that the accretion of primordial gas from the surroundings is more efficient in more massive systems , given their higher cross section per unit mass ( see pm04 ) .
the model galaxy suffers a strong starburst which injects a large amount of energy into the interstellar medium , able to trigger a galactic wind , occurring at different times at different radii , mainly due to the radial variation of the potential well , which is shallower in the galactic outskirts .
after the onset of wind activity the star formation is assumed to stop and the galaxy evolves passively with continuous mass loss . in order to correctly evaluate the amount of energy driving the wind , a detailed treatment of stellar feedback is included in the code ( that takes into account the stellar lifetimes ) . in particular , the energy restored to the interstellar medium by both type - ia and
type - ii supernovae has been calculated in a self - consistent manner according to the time of explosion of each supernova and the characteristics of the ambient medium ( see pm04 for details ) .
the potential well that keeps the gas bound to the galaxy is assumed to be dominated by a diffuse and massive halo of dark matter surrounding the galaxy . in the following we
adopt the standard star formation rate @xmath6 before the onset of the galactic wind ( @xmath7 ) , where @xmath8 is the gas density , @xmath9 the star - formation efficiency ; otherwise we assume that @xmath10 .
we recall here that the adopted star - formation efficiency is @xmath11 , while the infall timescale is @xmath12 in the galactic core and @xmath13 at one effective radius ( of the diffuse light , @xmath14 ) , respectively .
these values were chosen by pm04 in order to reproduce the majority of the chemical and photometric properties of ellipticals such as : the @xmath15[mg / fe]@xmath16 ( e.g. * ? ? ?
* ) , the color - magnitude ( e.g. * ? ? ?
* ) , the mass - metallicity relation ( e.g. * ? ? ?
* ) as well as the observed gradients in metallicity ( e.g. * ? ? ?
* ) , @xmath15[mg / fe]@xmath17 ( e.g. * ? ? ?
* ) , and color ( e.g. * ? ? ?
@xcite recently extended this model to explain also the properties of hot x - ray emitting halos surrounding more massive spheroids .
the formation rate of gcs , @xmath18 , in the _ _ i-__th shell is assumed to be directly linked to its star formation rate @xmath19 via a suitable function of time _
t _ , radius @xmath5 , and metallicity _ z _ , which represents some scaling law between star formation rate @xmath20 and the star cluster formation @xmath21 and can be regarded as a gc formation efficiency . a similar relation between the average star formation rate per surface area and the star cluster formation
was recently found by @xcite to hold in nearby spiral galaxies .
in addition , the efficiency of cluster formation in massive ellipticals appears to be constant , where the mass ratio between the mass in star clusters and the baryons locked in field stars+gas is @xmath22 @xcite .
here we extend this surface density relation to 3-d space .
moreover , pmc06 showed that at a given galactocentric radius model galaxies are made of a csp , namely a mixture of several simple stellar populations ( ssps ) each with a single age and chemical composition .
the csp reflects the chemical enrichment history of the entire system , weighted by the star formation rate .
we define the stellar metallicity distribution @xmath23 as the distribution of stars belonging to a given csp as a function of [ z / h ] .
we can then write the globular - cluster metallicity distribution @xmath24 at a given radius @xmath5 and time @xmath25 as : @xmath26 where @xmath27 includes all the information pertaining to the connection between @xmath18 and @xmath20 . it is not trivial , and beyond the scope of the paper , to find an explicit definition for @xmath28 , which basically carries the information on the internal physics of gas clouds where gcs are expected to form . in the following
we will show that , for a few and sensible choices of @xmath27 , the observed multi - modality in the color distribution of globular clusters may be driven by the radial variations in the stellar population mix of ellipticals .
we will first adopt a constant function @xmath27 ( see sec . [ bimod ] ) and then allow @xmath27 to mildly vary with @xmath29 ( see sec .
[ ref : metals ] ) .
no absolute values for @xmath27 will be given , since our formalism deals with normalized distributions .
in particular , the _ total _ @xmath24 summed over all radial shells can be written as : @xmath30 similar equations hold for other gc distributions as a function of either [ mg / fe ] or [ fe / h ] . at this stage
it is useful to recall that @xmath31 can be represented in two following ways : @xmath32 as the fraction of mass of a csp which is locked in stars at any given metallicity @xcite . in the following we refer to this stellar metallicity distribution as @xmath33 ( msmd : mass - weighted stellar metallicity distribution ) ; @xmath34 as a fraction of luminosity of the csp in each metallicity bin .
this definition is closer to the measurement as it can be directly compared to the luminosity - weighted mean @xmath35 ( lsmd : luminosity - weighted stellar metallicity distribution ) at a given radius ( see * ? ? ?
* ; * ? ? ?
this classification is important since pmc06 showed that the @xmath33 and @xmath35 might differ , especially at large radii , even for old stellar populations .
the advantage of gcss , for which accurate ages are known , is that they directly probe the mass - weighted distributions . at this point
it is useful to recall that our the adopted chemical evolution model divides the galaxy in several non - interacting shells . in each shell
the time at which the galactic wind occurs is self - consistently evaluated from the _ local _ condition .
in particular , we follow martinelli et al .
( 1998 ) suggestion that gradients can arise as a consequence of a more prolonged sf , and thus stronger chemical enrichment , in the inner zones . in the galactic core ,
in fact , the potential well is deeper and the supernovae ( sne ) driven wind develops later relative to the most external regions .
this particular formation scenario leaves a characteristic imprint on the shape of both @xmath33 and @xmath35 and here we give some general considerations .
in particular , we can explain the slow rise in the low metallicity tail of the distributions as the effect of the initially infalling gas , whereas the onset of the galactic wind sets the maximum metallicity of the @xmath33 and @xmath35 .
in general , the suggested outside - in formation process reflects in a more asymmetric stellar metallicity distribution at larger radii , where the galactic wind occurs earlier ( i.e. closer to the peak of the star formation rate ) , with respect to the galactic center . the _ qualitative _ agreement between these model predictions and the observed stellar metallicity distributions derived at different radii by harris & harris ( 2002 , see their fig .
18 ) for the stars in the elliptical galaxy ngc 5128 is remarkable .
if confirmed from observations in other ellipticals , the expected sharp truncation of @xmath33 at large radii might be the first direct evidence of a sudden and strong wind which stopped the star formation earlier in the galactic outskirts ( see pmc06 and pipino , dercole , & matteucci , in preparation ) .
the general presence of multi - modal gcss implies that their host galaxies did not form in a single , isolated monolithic event , but experienced spatially and/or temporally separated star - formation bursts . in the recent past , both semi - analytic and hydrodynamic simulations of galaxy formation attempted to follow the process of gc formation @xcite , but neither could produce a clearly bimodal mdf in their simulated gcss .
in this section we will show how to obtain a bimodal metallicity distribution function for gcs @xmath36 starting from single - mode stellar metallicity distribution functions @xmath31 ( commonly known as g - dwarf - like diagrams ) for the csp inhabiting different radii of a prototypical elliptical galaxy according to equation [ gcmd1 ] . as stressed in the introduction ,
the multi - modality in gcss varies as a function of host galaxy properties ( e.g. mass , morphological type , etc . ) . here
we try to match the distributions resulting from the recent compilation of spectroscopic data by puzia et al .
( 2006 , hereafter p06 ) , which samples the typical bimodal color distribution of gcs in nearby galaxies ( see also * ? ? ?
* ; * ? ? ?
this is illustrated in figure [ colsamp ] , where we plot the @xmath37 color distribution of the p06 sample of gcs in elliptical galaxies ( _ top panel _ ) together with those of gcs in ngc 4472 and the milky way ( _ middle and bottom panel _ ) .
ngc 4472 is the most luminous elliptical in the virgo galaxy cluster and hosts a gc system with a prototypical color bimodality ( e.g. * ? ? ?
* ) . to allow direct comparison with the p06 sample , we use gcs in ngc 4472 that are brighter than @xmath38 since the p06 sample includes only the brightest gcs in nearby early - type galaxies in order to maximize the s / n of their spectra
the resulting color distribution is remarkably similar to the one of the p06 sample , which assures that the p06 sample includes a representative sampling of the gc color bimodality in massive elliptical galaxies .
however , the comparison with the milky way gcs shows that the p06 sample covers few of the most metal - poor gcs .
therefore , the _ bimodality _ which we refer to in the following may not be the same as the one observed in spirals or in some elliptical galaxies , where a substantial population of metal - poor clusters with [ z / h ] @xmath39 is present ( e.g. * ? ? ?
we do not include the dynamical evolution of gcs in our model , since we are considering only the brightest ( most massive , i.e. @xmath40 , see also puzia et al .
2004 , a&a 415 , 123 ) clusters in nearby galaxies as comparison sample .
the comparison sample includes gcs much brighter than the typical turnover magnitude of the globular cluster luminosity function and we , therefore , do not expect significant differential dynamical evolution for these massive systems ( see * ? ? ? * for details ) .
in fact , it has been shown ( e.g. fall & zhang , 2001 ) that the timescales for both evaporation by two body relaxation and tidal stripping of star clusters is longer than a hubble time for gcs more massive than @xmath41 . in our model
, the number of clusters formed in a star - formation burst of a given strength is adjusted to match the observations .
hence , the absolute scaling of gc numbers is arbitrary , i.e. , within physical limitations of the star formation rate any number of gcs can be reproduced by adjusting the function @xmath27 in equation [ gcmd1 ] .
if , however , metal - poor and metal - rich gcs are on systematically different orbits and experience significantly different dynamical evolutions the effect of tidal disruption might be slowly changing relative gc numbers with time .
another complication is the variation of the initial star - cluster mass function , in particular as a function of metallicity .
modeling these effects requires detailed knowledge of the orbital characteristics and chemo - dynamical processes that lead to star cluster formation , and goes far beyond the scope of this work .
we keep these potential systematics in mind , but expect negligible impact on our analysis . in order to make a first - order comparison between our model predictions and the observed @xmath36 at @xmath42 gyr
, we first focus on the simple case in which : @xmath43 with @xmath44 and @xmath45 , @xmath46 . the first term ( @xmath47 )
corresponds to a population typical of the galaxy core ( well inside @xmath48 ) .
the second term ( @xmath49 ) represents @xmath23 in the outer regions . in order to take into account the different amounts of stars formed in each galactic region ,
we point out that the stellar metallicity distributions entering equation [ gcmd1 ] were not normalized . as a first step ,
we take several values for the weights @xmath50 in order to mimic different mixtures of the two gc populations .
in particular , we used the relative numbers of the _ red _ ( here identified as the metal - rich core population ) and the _ blue _ globular clusters ( the _ halo _ metal - poor population ) as a function of galactocentric radius for the elliptical galaxy ngc 1399 @xcite .
our particular choice is driven by observationally motivated values for the weights , although we realize that ngc 1399 is a quite peculiar , massive cd elliptical and might not be representative of less massive systems . in the context of this first step , the weights might reflect the effects of the projection on the sky of a three - dimensional structure .
however , we show below that the results do not strongly depend on the weights .
therefore they might be interpreted as mean values and could be changed if one decides to model a particular galaxy , with a different ellipticity , inclination , and luminosity profile . in figure [ gd_mass ]
we show the globular - cluster metallicity distribution @xmath24 by mass ( computation based on @xmath33 ) in two radial bins for three particular choices of weights .
in particular in the following we will use the ratios @xmath51 , @xmath52 and @xmath53 in order to define the theoretical _ innermost , intermediate and outermost _ sub - sample of the gcs , respectively .
these @xmath24 will be compared against subsamples of the p06 data , obtained by selecting gcs with either @xmath54 ( in the case of the innermost population ) or @xmath55 ( for the intermediate and the outemost cases , respectively ) , unless otherwise stated . in order to plot the different cases on the same scale
we normalize each @xmath24 by its maximum value . in the left panel of figure [ gd_mass ]
the shaded histogram represents the innermost population .
our predictions match the data very well , especially in the metal - rich slope and the mean of the distribution .
the same happens for the _ pure core _ populations , which shows how the gcs might be used to probe the csp in ellipticals .
it should be remarked that a second peak centered at super - solar metallicity appears in the distribution predicted by our models , although not evident in the data of the particular radial sub - sample .
the right panel of figure [ gd_mass ] illustrates model predictions which are more representative of the galaxy as a whole ( either at @xmath56 , i.e. the _ intermediate _ population , or at several effective radii , the _ outermost _ population ) , and we consider them as the fiducial case .
these two cases look quite similar to each other and have clear signs of bimodality in remarkable agreement with the spectroscopic data ( solid empty histogram , sub - sample of the p06 data with @xmath57 ) .
a kolmogorov - smirnov test returns @xmath58 probability that both model predictions and observations are drawn from the same parent distribution in the left panel of figure [ gd_mass ] .
the right panel statistics gives a lower likelihood of @xmath59 that both distributions have the same origin , which is mainly due to the observed excess of metal - poor gcs at large galactocentric radii compared to the model predictions .
the prediction of a super - solar metallicity globular cluster sub - population is entirely new and a result of the radially varying and violent formation of the parent galaxy . moving to the low - metallicity tail , we predict slightly fewer low metallicity objects than expected from observations .
but we recall the systematics mentioned in section [ compsample ] . in figure [ gd_mass2 ] ( left panel )
we show the results for a _ pure core _
gcs , namely one in which we adopt @xmath60 . in this quite extreme case
the observed gcs have been selected with radius @xmath61 .
the histogram reflects the shape of a g - dwarf - like diagram expected for a typical csp inhabiting the galactic core .
this finding is particularly important , because it might offer the opportunity to resolve the ssps in ellipticals , at variance with data coming from the integrated spectra which deal with luminosity - weighted quantities .
whereas in figure [ gd_mass2 ] ( right panel ) , the intermediate population is compared to a sub - sample of p06 gcs with @xmath62 .
this is to show that the multimodality is not an artifact due to the particular radial binning adopted in this paper .
figure [ gd_lum ] shows the @xmath63-band luminosity - weighted @xmath64 for which the computation is based on @xmath35 .
this metallicity distribution has been obtained by converting the mass in each [ z / h ] bin of the previous figure into @xmath65 , by means of the @xmath66 ratio computed by @xcite as a function of [ z / h ] for 13 gyr old ssps . due to the well - known increase of the @xmath66 in the high metallicity tail of the distribution and @xmath35 predicted for the same csp . ] , we notice in figure [ gd_lum ] that now the second peak has a smaller intensity in all the cases .
the corresponding diffuse - light population goes undetected in integrated - light studies . in any case , the conclusions reached by analyzing figure [ gd_mass ] are not significantly altered .
we conclude that our analysis is not significantly biased by some metallicity effect which may alter the shape of the observed @xmath36 by luminosity .
we stress the power of gcss in disentangling stellar sub - populations in massive ellipticals , due to their nature as simple stellar populations that can be directly compared to ssp model predictions , unlike diffuse light measurements .
even with this simple parametrization , where @xmath27 in equation [ gcmd2 ] does not depend on metallicity , we suggest that the bimodality for the metal - rich gcs is the result of different shapes of @xmath33 ( and the @xmath35 ) at different galactocentric radii .
finally , in figure [ gd_mgfe ] we show the [ mg / fe ] distributions for gcs divided in radial bins as in figure [ gd_mass ] . according to pmc06 ,
these [ mg / fe ] distributions are narrower , more symmetric , and exhibit a smaller radial variation with respect to the [ z / h ] distributions . in any case ,
a small degree of bimodality is still present .
we point out the impressive agreement with the spectroscopic observations by p06 .
there is a rather large discrepancy between data and models at the high-[mg / fe ] end .
hence , the corresponding kolmogorov - smirnov likelihood tests return a probaility of 1% ( for inner sample ) and 95% ( for the outer ones ) .
if we limit the model predictions to [ mg / fe ] @xmath67 0.8 dex , the agreement slightly improves , reaching a 10% probablity in the inner region .
however , the inner field data still does not reach the extreme [ mg / fe ] values of the models .
in fact , due to the monotonic decrease of the [ mg / fe ] as a function of either metallicity or time ( see pm04 ) , the lack of low - metallicity gcs , evident from figure [ gd_mass ] , translates into a lack of @xmath0-enhanced clusters .
the [ mg / fe ] bimodality of our model predictions and the match with the spectroscopic measurements strongly imply that globular clusters in massive elliptical galaxies form on two different timescales .
their chemical compositions are consistent with an _ early _ mode with a duration of @xmath68 myr and a _ normal _ formation that lasted for @xmath69 myr . in fact , according to the time - delay model ( see matteucci , 2001 ) and given the typical star formation history of our model ellipticals , the [ mg / fe ] ratio in the gas - out of which the gcs form - is quickly and continuously decreasing with time .
we predict that the [ mg / fe ] ratio can be higher than 0.65 dex ( i.e. in the bins in which our predictions exhibit a deficit of gcs with respect to the observed distribution ) only in the first @xmath70 myr ( see also fig.3 in p06 and related discussion ) .
in fact , such a high value for the [ mg / fe ] can be attained only if very massive type ii sne contribute to the chemical evolution , without any contribution from either lower - mass type ii or type ia sne .
normal _ formation , instead , is the one already plotted in fig .
[ gd_mgfe ] and forms on a typical timescale of @xmath71 gyr .
more quantitatively , our initial theoretical gcmd predicts that only @xmath72 of the gcs forms at [ mg / fe ] larger than 0.65 dex . in order to improve the agreement with observations we require that the above fraction should be increased to @xmath73 . since star and globular cluster formation are expected to be closely linked ( e.g. *
* ) the same must be true for the diffuse stellar population of such galaxies .
we therefore foresee the presence of a similar [ mg / fe ] bimodality in the diffuse light of massive elliptical galaxies . unfortunately , there are not direct observations confirming our suggestions , until the metallicity distributions for the diffuse stellar component in ellipticals will become available for a number of galaxies .
indeed , the detection of bimodality in the [ mg / fe]-distribution might be a benchmark test for our predictions .
we will still refer to multiple gc sub - populations .
however , their differences ought to be ascribed only to the fact that they are created during an extended ( and intense ) star formation event during which the variation in chemical evolution is not negligible .
the radial differences originate from the fact that the galactic wind epoch is tightly linked to the potential , occurring later in the innermost regions ( e.g. * ? ? ?
* ; * ? ? ?
moreover , pm04 and pmc06 found that also the infall timescale is linked to the galactocentric radius .
in particular , it lasts longer in the more internal regions , owing to the continuous gas flows in the center of the galactic potential well .
in particular , we recall that in our model the core experiences a longer ( @xmath74 gyr ) star - formation history with respect to the outskirts where the typical star - formation timescale is @xmath75 gyr . according to our models
the metal - rich population of gcs in massive elliptical galaxies may consist of multiple sub - populations which basically play the same role as the csps populating each galactocentric shell in our framework of the global galaxy evolution . at the same time , we point out the lack of our models to produce a significant fraction of metal - poor gcs similar to the halo gc population in the milky way , with the caveat that the star formation histories are very different .
this , in turn , suggests that gcss in giant elliptical galaxies were assembled by accretion of a significant number of metal - poor gcs . in this section
we explore the approach outlined in equation [ gcmd1 ] , by introducing the effect of metallicity in the function _
f_. in particular , we start by assuming that @xmath76}\!<\!-1)}{f(t , r_i,{\rm [ z / h]}\!>\!-1)}=2 \ ; , \ ] ] roughly following what was found for the gcs of the most nearby giant elliptical galaxy ngc 5128 ( centaurus a ) by @xcite ( @xcite , hereafter hh02 ) for their `` inner field '' , regardless of the radius of the @xmath77-th shell . since the final distributions are normalized , the actual zeropoint of the function @xmath27 is not relevant .
our model predictions are plotted in figure [ gc_hh ] .
we notice a modest increase of the low - metalliticy tail of the distribution with respect to the simple picture sketched in section [ bimod ] without metallicity dependence , as well as a lower fraction of globular clusters populating the high - metallicity peak . including the metallicity dependence leads to an ambiguous change in agreement with the general observed trend .
hence , no firm conclusions on the real need for a metallicity dependence can be drawn .
similar results are obtained in the more realistic case in which _ f _ is a linearly decreasing function of [ z / h ] . at variance with hh02 , we chose to adopt the same scaling irrespective of galactocentric radius , for the following reason . despite the fact that the hh02 stellar metallicity distributions @xmath23 as functions of [ z / h ] confirm both the shape and the radial behavior of our model predictions for the mass - weighted stellar metallicity distribution @xmath33 ( compare their fig . 7 with pmc06 fig .
4 ) , care should be taken when comparing their results for @xmath23 as a function of [ fe / h ] .
the latter , in fact , had been obtained by assuming a particular trend in the [ @xmath0/fe ] as a function of galactocentric radius which disagrees with the results of our detailed chemical evolution model .
in particular , we find an offset of at least 0.2 dex in the sense that [ fe / h ] @xmath78 [ fe / h ] @xmath79 at a given metallicity ( [ z / h ] ) .
this disagreement becomes larger either at very low metallicity or at larger galactocentric radii , where we expect a stronger @xmath0-enhancement .
once the pm04 value for [ fe / h ] is adopted in fig .
18 of hh02 , we find that : @xmath32 for the inner halo , the stellar @xmath33 should be shifted by @xmath75 dex toward lower metallicities , removing any metallicity effect , and @xmath34 for the outer halo the discrepancy between the stellar @xmath33 and the @xmath36 should be reduced .
nevertheless , we believe that some decrease with time of the function _ f _ could be motivated by theoretical arguments .
in fact , recent work ( e.g. * ? ? ? * ; * ? ? ?
* ) shows that gcs of all ages preferentially form in turbulent high - pressure regions .
if we interpret the decrease in the efficiency of star formation ( inside the gas clouds that form gcs ) , as a function of the ambient pressure @xcite as a proxy for the temporal behaviour of our function _
f _ , we find again a reduction of a factor @xmath80 from the early high - pressure epochs to a late , more quiescent evolutionary phase . for our fiducial model
we predict a ratio of metal - poor ( namely with [ z / h ] @xmath81 ) to metal - rich gcs of @xmath75 .
previous photometric surveys found that the typical value for gc systems in elliptical galaxies is close to unity @xcite .
provided a linear color - metallicity transformation ( see also section [ othermechs ] ) , a possible explanation for the discrepancy between our models and the observations might be obtained by boosting the metal - poor population by a factor of @xmath82}\!<\!-1)}/{f(t , r_i,{\rm [ z / h]}\!>\!-1)}\ge 5 $ ] .
another way to solve the discrepancy is to assume that all the _ missing _ globular clusters have been accreted from the surroundings , e.g. from dwarf satellites ( e.g. * ? ? ?
we estimate the amount of the accreted metal - poor gcs , needed to achieve a ratio close to 1 , as a factor of @xmath83 of the number of globular clusters initially formed inside the galaxy .
a natural consequence of the scenario depicted in sections [ bimod ] and [ ref : metals ] is that , at a given galactocentric radius , the mean metallicity and [ @xmath0/fe ] ratios of a gcs coincide with the mass - weighted [ @xmath15z / h@xmath84 and [ @xmath85/fe@xmath84 of the underlying stellar population , because the gc quantities are calculated either from @xmath33 or the @xmath35 ( see eqs . 1 and 2 of pmc06 ) ,
unless the scaling function _
f _ is allowed to strongly vary with time .
we expect this to happen at least in the innermost gc sub - populations , in which the effects of the accretion of gcs from the environment can be reasonably neglected .
in particular , pm04 predict that more massive galaxies should show higher [ @xmath85/fe@xmath84 and [ @xmath15z / h@xmath84 . if accretion plays a negligible role
, we expect the same correlations for the _ total _ gc population with host galaxy mass for the most massive systems , in agreement with current observations ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
in fact , if we perform the same study of the above sections for a @xmath86 galaxy ( see table 2 of pm04 for its properties ) , both peaks in @xmath36 shift their positions by about 0.2 dex to higher [ z / h ] .
this is in good agreement with the results of ( * ? ? ?
* see their figure 13 and 14 ) .
this trend holds for smaller objects as well .
if we apply the procedure to a @xmath87 galaxy ( model iib of pm04 ) , we find that the metal - rich peak shifts towards a lower metallicity by 0.3 dex ( with respect to our fiducial model with @xmath88 ) , while the other peak is now centered around [ z / h ] @xmath89 dex .
in particular , we find a faster decrease in the mean metalliticy of the metal - poor gcs than for the metal - rich ones , again in agreement with the @xcite results .
interestingly , the ratio of metal - poor to metal - rich cluster increases up to @xmath90 for less massive halos .
we recall that in the pm04 scenario , the low - mass galaxies are those forming on a longer timescales and with a slower infall rate .
therefore , we suggest that the combination of these factors is likely to at least partly explain the change of the gc distributions in different galaxy morphologies .
this is especially the case in dwarf galaxies , where star formation is slow and still on - going , together with the fact that the probability for a substantial change in the pressure of the interstellar medium relative to its initial values is higher than in ellipticals , thus implying a much stronger variation of @xmath27 with time .
it has been suggested that gc populations are produced during major merger events which would lead to present - day ellipticals and their rich gcss ( e.g. * ? ? ?
* ; * ? ? ?
subsequent studies ( e.g. * ? ? ?
* ; * ? ? ?
* ) challenged this view by pointing out the much higher @xmath91 and more metal - rich gcss in early - type galaxies compared to those of spiral and irregular galaxies , which are thought to represent the early building blocks of massive ellipticals . in the following ,
we study the impact of the merger hypothesis on the predictions of our simulations . in order to do that
, we extended the procedure sketched in the previous sections to the _ merger models _ presented by @xcite ( @xcite , hereafter pm06 ) . in this paper , the effects of late gas accretion episodes and subsequent merger - induced starbursts on the photo - chemical evolution of elliptical galaxies have been studied and compared to the picture of galaxy formation emerging from pm04 ; in particular the pm04 best model is taken here as a reference . by means of the comparison with the colour - magnitude relations and the @xmath92mg / fe@xmath93$]-@xmath94 relation observed in ellipticals ( e.g. * ? ? ?
* ) , pm06 showed that either bursts involving a gas mass comparable to the mass already transformed into stars during the first episode of star formation and occurring at any redshift ( major mergers ) , or bursts occurring at low redshift ( i.e. @xmath95 ) and with a large range of accreted mass ( minor mergers ) , are ruled out .
the reason lies in the fact that the chemical abundances in the ism after the galactic wind ( and before the occurrence of the merger ) are dominated by type ia sn explosions , which continuously enrich the gas with their ejecta ( mainly fe ) .
when the merger - induced starburst occurs , most stars form out of this enriched gas ( thus , e.g. , lowering the total @xmath92mg / fe@xmath96 $ ] ) ; at the same time , we expect the metallicities of gcs formed out of this gas to be on average higher and their [ mg / fe ] ratios to be lower than those of the bulk of stars and gcs formed in the initial starburst . in this work
we present the case in which the galaxy accretes a gas mass @xmath97 at @xmath98 gyr ( i.e. @xmath99 gyr after the onset of the galactic wind ) .
we make this choice for several reasons : @xmath77 this model is quite similar to the pm06 models _ b _ and _ g _ , which were among those in good agreement with observations of the diffuse galaxy light properties .
@xmath100 ) the formation epoch of the bulk of these _ second generation _ gcs can not occur @xmath101 gyr later than @xmath102 , because the majority of gcs in the most massive elliptical galaxies studied today appear old within the age resolution of current photometric ( @xmath103 ) and spectroscopic studies ( @xmath104 ) .
finally , the composition of the newly accreted gas is assumed to be primordial ( see pm06 for a detailed discussion ) , but we remark that we reach roughly the same conclusion in the case of solar composition , in order to mimic some pre - enrichment for the newly accreted gas .
we point out that , lacking dynamics , pm06 presented their results for one - zone models .
therefore , in this section we are considering equation [ gcmd1 ] limited to only one shell .
in this way we can check whether the single merger hypothesis alone is enough to produce some bimodality in the total globular - cluster metallicity distribution @xmath36 , and if it is consistent with the predictions based on our fiducial model described in section [ ref : metals ] .
we show our results in figure [ gd_merg ] and [ gd_mgfe2 ] .
we notice a clear change in the overall shape of the metallicity distribution @xmath36 with respect to the cases shown in the previous sections , in the sense that now @xmath36 is narrower and dominated by objects with super - solar metallicity ( and sub - solar [ mg / fe ] ratios ) with a dominant population at [ z / h ] @xmath105 , which is not prominent in the observations of p06 .
the high - metallicity globular cluster populations dominate the metallicity distribution which is at variance with both the results from previous sections and the observations . our merger model does not include the accretion of globular clusters that were already formed within the accreted satellite galaxies .
the inclusion of this effect could remedy the match between models and observations at low metallicities , as the typical gc in a dwarf galaxy is metal - poor ( e.g. * ? ? ?
* ; * ? ? ?
* ) and their addition to the initial gc population would enhance the total number of metal - poor gcs and improve the fit to the data . however , these clusters need to be @xmath0-enhanced to match the observations .
the impact of gc accretion on our post - merger model predictions will be studied in detail in a future paper . here
, we remark that the time at which the purely gaseous subsequent merger event can occur ( which does not import already formed gcs ) , is limited by the onset of the galactic wind , after which the type - ia sne dominates the nucleosynthesis , and needs to be completed at @xmath106 gyr after the first starburst .
however , this time constraint implies that such merger events would overlap with the initial starburst and be mostly indistinguishable from each other .
such a scenario closely resembles the searle - zinn scenario @xcite , in which galaxy halos are formed from the agglomeration of gaseous protogalactic fragments . later
merger events are excluded in our models , as they would produce gcs with sub - solar [ mg / fe ] ratios which is at variance with the observations .
note also that the fraction of gcs at [ z / h]@xmath107 can be recovered in our models only if the cluster formation at low metallicity is enhanced ( e.g. using a value of 10 instead of 2 in eq.[metaldepnd ] ) .
however , even in the case in which we adopt some @xmath108 strongly declining with total metallicity , which may alter the shape of @xmath36 enhancing the low - metallicity tail and thus improving the agreement with observations , the position of both super - solar metallicity peaks will not change , remaining at variance with the data .
the picture emerging from our analysis is far from being the general solution to explaining the complexity of gc color distributions , and it suggests only a scheme in which multiple mechanisms could be at work together , either broadening or adding features to the observed distributions . for instance , @xcite suggested that the color bimodality could arise from the presence of hot horizontal - branch stars ( so far not accounted for in ssp model predictions ) that results in a non - linear color - metallicity transformation producing two color peaks from an originally single - peak metallicity distribution .
we tested this scenario on our fiducial model gc metallicity distribution , by applying to each ssp the following transformation from [ fe / h ] to the @xmath109 color : @xmath110 } + \gamma \,{\rm [ fe / h]}^2 \\\nonumber + \delta \,{\rm [ fe / h}]^3 + \epsilon \,{\rm [ fe / h]}^4 \label{eq_yoon}\end{aligned}\ ] ] the numerical values of the coefficients are given in table [ tab : coeff ] and the relation was adopted from @xcite and is consistent with the best - fit relation presented in their figure 1b .
we show our results in figure [ yoon ] .
cr @xmath0 & 1.5033 + @xmath111 & 0.172774 + @xmath112 & @xmath113 + @xmath114 & @xmath115 + @xmath116 & @xmath117 + since we start from symmetric metallicity distributions , the non - linear transformation seems to work and produce a color bimodality for the csp inhabiting the @xmath118 shell ( figure [ yoon ] , solid line ) , although the bimodality is slightly exaggerated compared to real data ( see figure [ colsamp ] ) . in fact , a look at the color distribution which we obtained for the sole @xmath119 shell reveals that it still has one peak and is roughly symmetric ( figure [ yoon ] , dashed line ) .
obviously , since the @xmath120 [ fe / h ] relationship is meant to explain the gcs color bimodality without invoking any other effect , we did not combine the two histograms , either according to eq.[gcmd1 ] or to eq .
[ gcmd2 ] in our models , as we are comparing metallicity distributions to spectroscopic measurements .
it is of great importance to investigate this transformation with large and homogeneous data sets that cover a wide enough metallicity range to allow a robust analysis of the non - linear inflection point in the color - metallicity transformation .
however , as a result of figure [ yoon ] , we point out that the color bimodality typically found for gcss in massive early - type galaxies might be only partly due to a non - linear color - metallicity transformation .
another effect put forward by , e.g.,@xcite is the claim that gcs might have undergone a self - enrichment phase at the early stages of their formation , and therefore some gcs could have experienced a boost in metallicity which would be not representative of the metallicity of their parent gas cloud . finally , as already mentioned above in section [ ln : merger ] , some gcs residing in the outermost regions of the galaxies ( e.g. * ?
* ) could have experienced entirely different chemical enrichment histories at the time of their formation and later been added to a more massive system through accretion ( e.g. * ? ? ?
the inclusion of these effects goes far beyond the scope of this work , but we remind the reader that all the aforementioned effects might influence the interpretation of any globular cluster color and metallicity distribution .
by means of the comparison between pm04 s best model predictions for the radial changes in the csp chemical properties and the recent spectroscopic data on the metallicity distributions of extragalactic gcss from @xcite , we are able to derive some conclusions on the gc metallicity distributions in massive elliptical galaxies . in particular , we focused on the main drivers of the multi - modality that is observed in the majority of gcss in massive elliptical galaxies .
our main conclusions are : * we show that the observed multi - modality in the gc metallicity distributions can be ascribed to the radial variation in the underlying stellar populations in giant elliptical galaxies
. in particular , the observed gcss are consistent with a linear combination of the gc sub - populations inhabiting different galactocentric radii projected on the sky . *
a new prediction of our models , which is in astonishing agreement with the spectroscopic observations , is the presence of a super - solar metallicity mode that seems to emerge in the most massive elliptical galaxies . in smaller objects , instead ,
this mode disappears quickly with decreasing stellar mass of the host galaxy . * our models
successfully reproduce the observed [ mg / fe ] bimodality in gcss of massive elliptical galaxies .
this , in turn , suggests a bimodality in formation timescales during the early formation epochs of gcs in massive galaxy halos .
the two modes are consistent with an early ( initial ) and later ( triggered ) formation mode .
* since the gc populations trace the properties of galactic csps in our scenario , we predict an increase of the mean metallicity of the cluster system with the host galaxy mass , which closely follows the mass - metallicity relation for ellipticals .
moreover , we expect that a major fraction of the gcs ( i.e. those born inside the galaxy ) follows an age - metalliticity relationship , in the sense that the older ones are also more @xmath0-enhanced and more metal - poor .
* the role of host galaxy metallicity in shaping the observed gc metallicity distribution is non - negligible , although its effects have been estimated to change the function @xmath121 by a factor of @xmath122 , in order to match the sample of @xcite .
either a non - linear color - metallicity transformation , or a stronger metallicity effect , and/or accretion of gcs from the surrounding environment is needed to explain a ratio of metal - poor to metal - rich gcs close to unity , as reported for ellipticals based on results from photometric surveys . * merger models which include the later accretion of primordial and/or solar - metallicity gas predict a shape for the gc metallicity distribution which is at variance with the spectroscopic observations .
we thank the referee for a careful reading of the paper .
warmly thanks s.recchi for useful discussions . a.p .
acknowledges support by the italian ministry for university under the cofin03 prot .
2003028039 scheme .
acknowledges support by nasa through grants go-10129 and go-10515 from the space telescope science institute , which is operated by aura , inc . , under nasa contract nas5 - 26555 , and the support in form of a plaskett research fellowship at the herzberg institute of astrophysics .
arimoto , n. , @xmath123 yoshii , y. 1987 , a@xmath123a , 173 , 23 ashman , k. m. , & zepf , s. e. 1992 , , 384 , 50 ashman , k. m. & zepf , s.e .
1998 , globular cluster systems , cambridge university press beasley , m. a. , baugh , c. m. , forbes , d. a. , sharples , r. m. , & frenk , c. s. 2002 , , 333 , 383 bower , r. g. , lucey , j. r. , & ellis , r. s. 1992 , mnras , 254 , 601 bressan , a. , chiosi , c. , fagotto , f. 1994 , apjs , 94 , 63 brodie , j.p . , & huchra , j.p . , 1991 , apj , 379 , 157 brodie , j.p . , & strader , j. , 2006 , astro - ph/0602601 carollo , c.m . , danziger , i.j . , @xmath123 buson , l. 1993 , mnras , 265 , 553 chandar , r. , fall , s. m. , & whitmore , b. c. 2006 , , 650 , l111 c^ ot ' e , p. , marzke , r. o. , & west , m. j. 1998 , , 501 , 554 dirsch , b. , richtler , t. , geisler , d. , forte , j.c .
, bassino , l.p . ,
& gieren , w.p . , 2003 , apj , 125 , 1908 dirsch , b. , schuberth , y. , & richtler , t. 2005 , , 433 , 43 elmegreen , b. c. 2004 , asp conf .
ser . 322 : the formation and evolution of massive young star clusters , 322 , 277 elmegreen , b. g. , & efremov , y. n. 1997 , , 480 , 235 elmegreen , b. g. , & scalo , j. 2004 , , 42 , 211 faber , s.m . , worthey , g. , @xmath123 gonzalez , j.j .
1992 , in iau symp .
n.149 , eds .
b. barbuy @xmath123 a. renzini , p. 255
fall , s.m .
, & zhang , q. 2001 , , 561 , 751 forbes , d. a. , brodie , j. p. , & grillmair , c. j. 1997 , , 113 , 1652 gallazzi , a. , charlot , s. , brinchmann , j. , white , s. d. m. , & tremonti , c. a. 2005 , mnras , 362 , 41 gebhardt , k. , & kissler - patig , m. 1999 , , 118 , 1526 gibson , b.k . , 1996 , mnras , 278 , 829 gnedin , o. y. , & ostriker , j. p. 1997
, , 474 , 223 greggio , l. , 1997 , mnras , 285 , 151 harris , w. e. 1991 , , 29 , 543 harris , w. e. 1996 , , 112 , 1487 harris , w. e. 2001 , in saas - fee advanced school on star clusters , ed .
l. labhardt & b. binggeli ( course 28 ) , springer , new york harris , w. e. , harris , g. l. h. , & mclaughlin , d. e. 1998 , , 115 , 1801 harris , w. e. , & harris , g. l. h. 2002 , aj , 123 , 3108 holtzman , j. a. , et al .
1992 , , 103 , 691 kissler - patig , m. , brodie , j. p. , schroder , l. l. , forbes , d. a. , grillmair , c. j. , & huchra , j. p. 1998a , , 115 , 105 kissler - patig , m. , forbes , d. a. , & minniti , d. 1998b , , 298 , 1123 kravtsov , a. v. , & gnedin , o. y. 2005 , , 623 , 650 kundu , a. , & whitmore , b. c. 2001a , , 121 , 2950 kundu , a. , & whitmore , b. c. 2001b , , 122 , 1251 larsen , s. s. , brodie , j. p. , huchra , j. p. , forbes , d. a. , & grillmair , c. j. 2001 , , 121 , 2974 larsen , s. s. , & richtler , t. 2000 , a&a , 354 , 836 lee , j .- w .
, lpez - morales , m. , & carney , b. w. 2006 , , 646 , l119 li , y. , mac low , m .- m . , &
klessen , r. s. 2004 , , 614 , l29 lotz , j. m. , miller , b. w. , & ferguson , h. c. 2004 , , 613 , 262 lutz , d. 1991 , , 245 , 31 maraston , c. 2005 , , 362 , 799 martinelli , a. , matteucci , f. , colafrancesco , s. , 1998 , mnras , 298 , 42 matteucci , f. 2001 , the chemical evolution of the galaxy , kluwer academic publishers , dordrecht mclaughlin , d. e. 1999 , , 117 , 2398 mendez , r.h . , thomas , d. , saglia , r.p . ,
maraston , c. , kudritzki , r.p . , & bender , r. , 2005 , apj , 627 , 767 nomoto , k. , hashimoto , m. , tsujimoto , t. , thielemann , f.k .
, kishimoto , n. , kubo , y. , nakasato , n. , 1997 , nuclear physics a , a621 , 467 osterbrock , d. e. , & ferland , g. j. 2006 , astrophysics of gaseous nebulae and active galactic nuclei , 2nd . ed . by d.e .
osterbrock and g.j .
ferland .
sausalito , ca : university science books , 2006 , pagel , b. e. j. , @xmath123 patchett , b.e . 1975 , mnras , 172 , 13 peletier , r.f . , davies , r.l . ,
illingworth , g.d . ,
davis , l.e .
, cawson , m. 1990 , aj , 100 , 1091 peng , e. w. , et al . 2006 , , 639 , 95 pipino , a. , matteucci , f. 2004 , mnras , 347 , 968 ( pm04 ) pipino , a. , kawata , d. , gibson , b. k. , & matteucci , f. 2005 , , 434 , 553 pipino , a. , matteucci , f. , & chiappini , c. 2006 , , 638 , 739 ( pmc06 ) pipino , a. , & matteucci , f. 2006 , , 365 , 1114 ( pm06 ) puzia , t. h. , kissler - patig , m. , brodie , j. p. , & huchra , j. p. 1999
, , 118 , 2734 puzia , t. h. , et al .
2004 , , 415 , 123 puzia , t. h. , kissler - patig , m. , thomas , d. , maraston , c. , saglia , r. p. , bender , r. , goudfrooij , p. , & hempel , m. 2005 , , 439 , 997 puzia , t. h. , kissler - patig , m. , & goudfrooij , p. 2006
, , 648 , 383 , ( p06 ) recchi , s. , & danziger , i. j. 2005 , , 436 , 145 renzini , a. 2006 , , 44 , 141 rhode , k. l. , & zepf , s. e. 2004 , , 127 , 302 salpeter , e. e. 1955 , apj , 121 , 161 schweizer , f. 1987 , nearly normal galaxies . from the planck time to the present , 18 searle , l. , & zinn , r. 1978 , , 225 , 357 sharina , m. e. , puzia , t. h. , & makarov , d. i. 2005 , , 442 , 85 thielemann , f. k. , nomoto , k. , hashimoto , m. 1996 , apj , 460 , 408 thomas , d. , maraston , c. , @xmath123 bender , r. , 2002 , ap@xmath123ss , 281 , 371 tinsley , b.m . , 1980 , apj , 241 , 41 van den hoek , l.b . , groenewegen , m.a.t .
1997 , a@xmath123as , 123 , 305 van den bergh , 1975 , ara&a , 13 , 217 weiss , a. , peletier , r.f . ,
matteucci , f. 1995 , a@xmath123a , 296 , 73 west , m. j. , ct , p. , marzke , r. o. , & jordn , a. 2004 , , 427 , 31 worthey , g. , faber , s.m . ,
@xmath123 gonzalez , j.j .
1992 , apj , 398 , 69 yoon , s .- j .
, yi , s. k. , & lee , y .- w .
2006 , science , 311 , 1129 zepf , s. e. , & ashman , k. m. 1993 , , 264 , 611 | the most massive elliptical galaxies show a prominent multi - modality in their globular cluster system color distributions .
understanding the mechanisms which lead to multiple globular cluster sub - populations is essential for a complete picture of massive galaxy formation . by assuming that globular cluster formation traces the total star formation and taking into account the radial variations in the composite stellar populations predicted by the pipino & matteucci ( 2004 ) multi - zone photo - chemical evolution code , we compute the distribution of globular cluster properties as a function of galactocentric radius . we compare our results to the spectroscopic measurements of globular clusters in nearby early - type galaxies by puzia et al .
( 2006 ) and show that the observed multi - modality in globular cluster systems of massive ellipticals can be , at least partly , ascribed to the radial variation in the mix of stellar populations .
our model predicts the presence of a super - metal - rich population of globular clusters in the most massive elliptical galaxies , which is in very good agreement with the spectroscopic observations .
the size of this high - metallicity population scales with galaxy mass , in the sense that more massive galaxies host larger such cluster populations .
we predict an increase of mean metallicity of the globular cluster systems with host galaxy mass , and forecast that those clusters that were formed within the initial galaxy halo closely follow an age - metallicity and @xmath0/fe - metallicity relation , where older clusters exhibit lower metallicities and higher @xmath0/fe ratios .
furthermore , we investigate the impact of other non - linear mechanisms that shape the metallicity distribution of globular cluster systems , in particular the role of merger - induced globular cluster formation and a non - linear color - metallicity transformation , and discuss their influence in the context of our model .
we find that a non - linear color - metallicity relation may be partly responsible for a color multi - modality . on the other hand
, the formation of globular clusters from subsequently accreted gas , either with primordial abundances or solar metallicity , delivers model predictions which are at variance with the observations , and suggests that a significant fraction of metal - poor globular clusters is accreted from satellite halos which formed their globular clusters independently . |
x - ray emission from accreting black holes probes the innermost relativistic regime .
emission lines produced in the inner accretion disk are expected to bear the imprints of the strong doppler shifts and gravitational red - shifts natural to this region .
advances in x - ray spectroscopy have made it possible to exploit relativistic iron lines as probes of black hole accretion and even measures of black hole spin ( for a review , see miller 2007 ) .
recently , relativistic line models have been developed in which black hole spin is a variable parameter that can be constrained through spectral fitting ( e.g. brenneman & reynolds 2007 ) ; this marks another advance .
_ suzaku _ is a major step forward for relativistic spectroscopy of stellar - mass black holes : its ccd cameras retain a high livetime fraction even at high flux levels , and its bandpass covers both the relativistic iron line region and more subtle disk reflection curvature expected at high energy .
the combination of new models and the capabilities of _ suzaku _ provide the opportunity to obtain robust spin constraints in a number of stellar - mass black holes .
gx 339@xmath04 is a recurrent black hole transient with a low - mass companion star .
its mass has been constrained to be greater than @xmath4 ( hynes et al .
the distance to gx 339@xmath04 is likely to be greater than 6 kpc and may be as great as 15 kpc ( hynes et al .
2004 ) ; a value of 8 kpc may be most likely ( zdziarski et al.2004 ) .
the radio jet observed in gx 339@xmath04 and its x - ray properties argue for an inner disk inclination that may be as low as @xmath5 ( gallo et al .
2004 , miller et al .
this inclination would imply a high black hole mass , but the inner disk and binary system need not have the same inclination .
prior observations of gx 339@xmath04 with _
chandra _ and _ xmm - newton _ have revealed lines suggestive of a black hole with a high spin parameter ( miller et al . 2004a , 2004b , 2006 ) .
observations of xte j1650@xmath0500 with _ xmm - newton _ and _ bepposax _ have also revealed skewed iron disk lines suggestive of high spin ( miller et al . 2002 ; miniutti , fabian , & miller 2004 ; rossi et al .
spectroscopy of gro j1655@xmath040 has sometimes revealed relativistic iron lines suggestive of near - maximal black hole spin ( miller et al.2005 ; diaz - trigo et al .
in contrast to these sources , iron lines observed in grs 1915@xmath6105 does not require a high spin parameter ( martocchia et al .
2002 ) .
in late 2006 , gx 339@xmath04 entered a new outburst with a slow rise phase ( swank et al .
2006 ) . in the spring of 2007
, we triggered an approved _ suzaku _ too observation .
rxte_/asm monitoring , we observed gx 339@xmath04 in an `` intermediate '' state ( for a review , see remillard & mcclintock 2006 ) . in the sections below
, we detail how we reduced and analyzed the spectra , and discuss the strengths and weaknesses of our measurements and methods .
we observed gx 339@xmath04 with _ suzaku _ on 2007 february 12 starting at 05:33:31 ( tt ) .
the nominal observation duration was 100 ksec , but different exposures were obtained due to observing constraints and the instrument modes chosen .
the xis pointing was used . in order to prevent photon pile - up
, the xis units were operated using the 1/4 window mode using a 0.3 second burst option . in general , xis on - source times of approximately 70 ksec were achieved .
this resulted in deadtime - corrected net 3x3 editing mode exposures of 10.3 ksec , 9.3 ksec , and 16.9 ksec for the xis0 , xis1 , and xis3 cameras , respectively .
the hxd / pin and hxd / gso cameras were operated in their default modes .
a net exposure time of 87.9 ksec was achieved using the hxd / pin .
we reduced version-2processed data .
it is important to note that the xis burst modes are not yet calibrated as well as more standard modes . as reproducibility is even more important when non - standard modes are used , we chose to use the xis and hxd pipeline spectra .
for the xis , these products include both source and background spectra from each camera ; for the hxd / pin , the pipeling products include only the source spectrum .
we generated xis redistribution matrix files ( rmfs ) and ancillary response files ( arfs ) using the tools `` xisrmfgen '' and `` xissimarfgen '' available in the heasoft version 6.4 reduction and analysis suite .
the `` xissimarfgen '' tool generates a response using a simulation ; we generated the recommended 400,000 photons in creating each ancillary response .
( xis spectra over - sample the true energy resolution of the instrument , as do the _ xmm - newton _ spectra treated in later sections .
though rmf files account for this , the reader should be aware that the resolution is over - sampled . ) to create an hxd / pin background file , we simulated non - x - ray background and cosmic x - ray backgrounds in xspec version 11.3 as per the instruction in the _ suzaku _ abc guide .
the standard pin response file for xis - nominal pointings was used .
the spectra from xis0 and xis1 matched each other closely , while the spectrum from xis3 has somewhat larger residuals ( in the si band , for instance ) . in order to focus on relativistic spectroscopy rather than calibration uncertainties , we restricted our analysis to xis0 and xis1 .
for all plausible continuum models , these spectra show strong residuals below 0.7 kev and above 10.0 kev ; we therefore restricted our xis fitting range to the 0.710.0 kev band .
similarly , the hxd / pin shows strong residuals below 12 kev .
the source is strong throughout the hxd / pin bandpass , so we considered the 12.070.0 kev band when fitting this spectrum .
all spectral fits reported in this work were made using xspec 11.3 .
errors reported in this work are 1@xmath7 errors obtained using the `` error '' command .
the high sensitivity of our _ suzaku _ spectra provides an unprecedented chance to study a relativistic disk reflection spectrum , and this is the focus of our analysis . especially given that the absolute flux calibration in our modes is still being refined , and given that contamination affects the effective area below 0.7 kev in ways that could affect disk continuum spectroscopy , focusing only on the residuals to fiducial continua is particularly pragmatic .
indeed , this is a general advantage of line and reflection spectroscopy over disk continuum spectroscopy .
figure 1 shows the _ suzaku _ spectra of gx 339@xmath04 fit with a simple model consisting of disk blackbody ( `` diskbb '' ) and power - law components , modified by interstellar absorption ( `` phabs '' ) .
the continuum is the best
fit continuum for each individual camera ( the parameters were not linked during the fit ) .
a gaussian emission line was added at 2.26 kev to account for an instrumental response defect .
the 4.0 - 7.0 kev and 15.0 - 40.0 kev ranges were ignored when fitting the data in order to show the disk reflection signatures clearly , and then restored when forming the data / model ratio shown in figure 1 ( as per miniutti et al .
. fits to the continuum with this simple model reveals a relativistic iron emission line ( miller et al .
2004a , 2004b , 2006 ) and a compton back - scattering hump .
the asymmetric shape of the broad line is exactly that predicted by relativistic disk line models , and the curvature of the compton reflection hump is clearly revealed in the hxd spectrum .
the broad - band relativistic disk spectrum is very similar to that revealed in _
spectra of the seyfert-1 agn mcg-6 - 30 - 15 ( reeves et al .
2006 , miniutti et al .
it must be noted that residuals remain in the 23 kev band .
these residuals are not physical , but rather related to aspects of the instrument response ( e.g. si and au edges ) that are not yet modeled correctly by the response files . owing to the quality of the spectra line and reflection spectra , we made fits with a physically - motivated model suited to constraining the spin of the black hole in gx 339@xmath04 .
the model included a disk blackbody component , a power - law component , and the constant density ionized disk ( cdid ) reflection model ( ballantyne , iwasawa , & fabian 2001 ) .
this disk reflection model includes line emission , and was convolved with the `` kerrdisk '' line function using the `` kerrconv '' model ( brenneman & reynolds 2006 ) .
we have chosed to use the `` kerrdisk '' model because radii are parameterized in terms of the innermost stable orbit for a given spin parameter ; other line models predict very similar profiles but differ in some details ( see dovciak , karas , & yaqoob 2004 and beckwith & done 2004 ) . in all fits ,
the outer reflection radius is assumed to be 400 times the innermost stable circular orbit .
the cdid reflection model is particularly well suited to high ionization regimes , and explicitly measures the inner disk ionization parameter ( @xmath8 ) .
the cdid model is an angle - averaged model , and does not include the disk inclination as a variable parameter .
this model provides a very good description of the spectra ( see figure 2 and table 1 ) , though the fit is not formally acceptable .
a high inner disk ionization parameter of log(@xmath9 ) = 4.1(2 ) is measured ; this is broadly consistent with prior results obtained in `` intermediate '' states ( e.g. miller et al .
the most important result obtained with this spectral model is the black hole spin parameter : @xmath10 .
gx 339@xmath04 is special in that excellent spectra have been obtained in each of the spectral states where relativistic lines are expected .
we next fit each of these spectra jointly to obtain the best possible constraint on the black hole spin parameter .
this procedure also ensures that state - specific changes ( such as disk ionization ) are not driving any spin constraints . using the same model
, we jointly fit our _ suzaku _ xis spectra , the `` very high '' state spectrum obtained with _ xmm - newton_/epic - pn camera , and the `` low / hard '' state spectra obtained with the _ xmm - newton _ mos1 and mos2 spectra in revolution 782 .
( the hxd spectrum considered above was omitted as its response is complex and drives up computation times . ) the same spectral and response files used in miller et al .
( 2004a ) and miller et al .
( 2006 ) were employed in this effort .
all spectra were fit in the 0.710.0 kev band except the low / hard state spectra , which were fit in the 0.79.0 kev band . for simplicity , an ionization parameter of @xmath11 was assumed for the intermediate state ( see above ) , and a value of @xmath12 was assumed for the low / hard state as per the results of modeling described in miller et al .
while each camera was again allowed to have its own continuum parameters , the spin parameter and inclination were determined by the joint fits .
this joint fit across states measures a spin parameter of @xmath13 .
this is very similar to the result obtained using the broad - band _ suzaku _ spectra of the intermediate state alone , and suggests that scattering in the disk atmosphere an effect expected to change across states can accurately be accounted for with good reflection models .
the very high state spectrum required an enhanced inner line emissivity ( @xmath14 within @xmath15 ; @xmath16 at larger radii ) ; this is common in the very high state ( miller et al.2002 , 2004a ; diaz trigo et al .
2007 ) . once again
, the overall fit is not statistically acceptable in a formal sense , but this is driven by narrow - band calibration problems .
we have analyzed _ suzaku _ spectra of the well - known recurrent black hole transient gx 339@xmath04 in outburst .
motivated by the quality of the broad - band spectra obtained , we fit the data with models capable of measuring the black hole spin parameter .
fits with a relativistically blurred disk reflection model give a spin parameter of @xmath17 .
as a check on this result , we jointly fit the _ suzaku _ xis spectra with ccd spectra in the `` very high '' and `` low / hard '' states obtained with _ xmm - newton_. this procedure yields a commensurate spin parameter of @xmath18 .
we therefore suggest that the best value of the spin parameter in gx 339@xmath04 is @xmath19 , where the systematic error aims to account for the error associated with considering a single state .
relativistic iron emission we do not regard the apparent changes in line - of - sight column density and inner disk radius in table 1 as robust .
the changes are more likely due in part to differences in flux calibration between _ xmm - newton _ and _ suzaku_. the apparent changes may also reflect real differences in disk ionization that falsely mimic radius varations ( merloni , fabian , & ross 2000 ) .
relativistic lines are also likely to be affected by such changes , especially compton scattering in the disk atmosphere .
disk reflection models attempt to account for these effects .
making joint fits to different states allows us to obtain a measure of the spin parameter in a way that is reasonably independent of changing scattering effects . to achieve a maximal spin parameter ,
a black hole must accrete approximately half of its mass ( volonteri et al . 2005 ) .
it is unlikely , then , that black holes in low - mass x - ray binary systems can achieve maximal spin through accretion alone .
the short lifespan of massive stars also makes it unlikely that black holes in high - mass binaries can achieve a maximal spin parameter through accretion .
factoring in these constraints and the range of black hole spins likely to be generated in supernova / grb events , gammie , shapiro , & mckinney ( 2004 ) suggest that stellar - mass black holes should have spins of @xmath20 .
the results we have obtained are consistent with this prediction .
accretion disks are necessarily an indirect measure of black hole spin , and the accuracy of all such measurements depends on how strong the contrast is between the inner edge of the accretion disk , and the plunging region within the isco .
the most advanced 3-d general relativistic mhd simulations of accretion disks are beginning to bear on this problem , though the case of a standard thin accretion disk has not yet been treated ( krolik & hawley 2002 ; krolik & hirose 2004 ) . the case of a thin disk has been treated simply by shafee , narayan , & mcclintock ( 2007 ) ; however , the most advanced and physically realistic treatment is given by reynolds & fabian ( 2008 ) .
the latter work makes use of sophisticated disk simulations and concludes that the contrast between the disk and plunging region is relatively high and that relativistic line diagnostics are robust .
this work partially validates the implicit assumption of a geometrically - thin disk with keplerian orbits .
although the relativistic modeling performed in this paper has permitted a significant advance over many prior efforts to characterize relativistic lines in stellar - mass black holes , our results must still be regarded as preliminary . in certain respects ,
disk reflection models are like stellar atmosphere models : the physics is known and the problems are tractable , but refinements can always be made .
the cdid model we have used is excellent and well suited to highly ionized disks ( ballantyne , iwasawa , & fabian 2001 ) , but astrophysical disks have a vertical density structure . moreover , averaging over viewing angle may act to wash - out important angle dependencies . at the time of writing , improved reflection models that assume vertical structure according to hydrostatic equilibrium and a hot thermal midplane spectrum are becoming available ( ross & fabian 2007 ; reis et al .
all current reflection models assume a single ionization zone ; however , the fact that only the inner radii contribute strongly to the red wing make this a robust simplification .
the use of relativistic lines to measure spin parameters in stellar - mass black holes holds enormous promise and some specific advantages over other techniques . to obtain spin constraints using the accretion disk continuum , the black hole mass , distance , and mass accretion rate
must be known .
the uncertainty on the mass can be as high as 30% ( remillard & mcclintock 2006 ) , and the accretion rate can be uncertain by a factor of a few . as an observational rather than empirical science , astronomy
is poorly suited to absolute flux ( luminosity ) measurements , and the difficulties with continuum spectroscopy are merely a special case of the general problem .
iron lines have the advantage of harnessing dynamical imprints to estimate the inner disk radius and black hole spin parameter .
a remaining difficulty for relativistic iron line measurements is that the geometry of the hard x - ray source illuminating the disk is unknown ; however , this is effectively encoded by the line emissivity index and can be determined observationally .
we thank charles gammie , laura brenneman , and randy ross for helpful discussions .
we thank the anonymous referee for helpful comments that improved this paper . this work has made use of the tools and services available through heasarc online service , which is operated by gsfc for nasa .
miller , j. m. , fabian , a. c. , nowak , m. a. , & lewin , w. h. g. l. , 2005 , in the proceedings of the tenth marcel grossmann meeting , eds .
m. novello , s. perez berliaffa , r. ruffini , singapore : world scientific lllllllllllll spectra & @xmath21 & @xmath22 & norm & & @xmath23 & @xmath24 & q & log(@xmath9 ) & @xmath25 & norm .
& @xmath26 + & ( @xmath27 ) & ( kev ) & ( @xmath28 ) & & ( @xmath29 ) & ( degrees ) & & & & & ( @xmath30 ) & + suzaku & 6.4(1 ) & 0.81(1 ) & 1.05(1 ) & & @xmath31 & @xmath32 & 3.0(1 ) & 4.1(2 ) & 1.92(4 ) & 0.57(2 ) & 2.5(1 ) & 11860/5076 + xmm vh & 5.1(1 ) & 0.80(1 ) & 1.63(1 ) & & @xmath33 & @xmath34 & 6.0(1 ) & 5.0(2 ) & 2.73(5 ) & 1.43(5 ) & 3.1(2 ) & 16190/7992 + suzaku is & 6.4(1 ) & 0.81(1 ) & 1.00(1 ) & & & & 3.00(6 ) & 4.0 & 1.99(3 ) & 1.22(2 ) & 2.4(1 ) & + xmm lh & 4.3(1 ) & 0.44(1 ) & 0.47(1 ) & & & & & 2.8 & 1.39(4 ) & 0.25(1 ) & 9.0(6 ) & | we report on a deep _ suzaku _ observation of the stellar - mass black hole gx 339@xmath04 in outburst . a clear , strong , relativistically - shaped iron emission line from the inner accretion disk is observed .
the broad - band disk reflection spectrum revealed is one of the most sensitive yet obtained from an accreting black hole .
we fit the _ suzaku _ spectra with a physically motivated disk reflection model , blurred by a new relativistic line function in which the black hole spin parameter is a variable .
this procedure yielded a black hole spin parameter of @xmath1 .
joint modeling of these _ suzaku _ spectra and prior _ xmm - newton _
spectra obtained in two different outburst phases yields a spin parameter of @xmath2 .
the degree of consistency between these results suggests that disk reflection models allow for spin measurements that are not strongly biased by scattering effects .
we suggest that the best value of the black hole spin parameter is @xmath3 .
although preliminary , these results represent the first direct measurement of non - zero spin in a stellar - mass black hole using relativistic line modeling .
[ firstpage ] |
the shape of the gamma - ray burst s ( grb s ) @xmath0 ms resolution lightcurves in the batse gamma - ray burst catalog @xcite carry an immense amount of information .
however , the chaging s / n ratio complicates the detailed comparative analysis of the lightcurves . during the morphological analysis of the grb
s @xcite a subclass with fast rise - exponential decay ( fred ) pulse shape were observed .
this shape is quite attractive because its fenomenological simplicity .
here we use a special wavelet transformation with a kernel function based on a fred - like pulse .
similar approach have been used by @xcite , but their base functions were constructed differently .
we have used the discrete wavelet transform ( dwt ) matrix formalism ( e.g. @xcite ) : here for an input data vector @xmath1 , the one step of the wavelet transform is a multiplication with a special matrix @xmath2 : @xmath3\ ] ] where the @xmath4 are the 4-stage fir filter parameters defining the wavelet . to obtain these values
we require the matrix @xmath2 to be orthogonal ( e.g. no information loss ) , and the output of the even ( derivating - like ) rows should disappear for a constant and for a fred - like @xmath5 input signal .
these requirements give two different solutions for @xmath4 : a rapidly oscillating one and a smooth one . in the following
we ll use the later one .
our filter process with the fred wavelet transform consists of the usual @xmath6 digital filtering steps . during the filtering we ll loose some information
, however this could be quite small . to demonstrate the efficiency of the algorithm on fig . 1 .
we reconstructed the 100 - 320 kev 64ms ligthcurve of batse trigger 0143 from the biggest 5% of the total wavelet coefficients .
the excellent reconstruction of each individual pulse is obvious .
the wavelet transformation algorithm divides the phase - space into equal area regions . on fig . 2 .
the wavelet transform are shown . here
the dark segments are the really important coefficients - however they cover only a small portion of the total area which explains the high efficiency of the reconstruction .
for a frequency - like wavelet scale analysis we would like to create a power - spectrum like distribution along the frequency axis
. however , one should be careful . in the classical signal
processing one uses the power spectrum from the fourier - transform , because the signals are electromagnetic - like usually , e.g. the power ( or energy ) is proportional to the square of the signal . here
the lightcurves measure photon counts so the signal s energy is simply the sum of the counts .
for this reason we approximate the signal s strength as a sum the magnitude of the coefficients along the given frequency rows .
this signal s strength indicate on fig 3 .
( batse trigger 0143 ) the maximum power to be around @xmath7 . in each energy channel
the signals are similar ( observe the logarithmic scale ) , because the signal is strong even at high energies ( channel 4 ) . for batse trigger 7343 ( with optical redshift @xmath8 ) one can observe a strong high frequency cutoff : some of the signal s high frequency part is missing .
however all the 4 channels are visible , while the maximum power is around @xmath9 .
it is interesting to remark that the signal s shape is quite similar to trigger 0143 if that is scaled down by a factor of @xmath10 in frequency .
the fred wavelet transform measures the similarity between the different wavelet kernel functions ( here all are fred - based ) and the actual signal . to quantify the similarity we define a magnitude cutoff in the wavelet space so , that the _ reconstructed _ @xmath11 value from the filtered data should be similar to the original values .
the @xmath11 value and its @xmath12 error from the photon count statistics could be easily determined from the original ligthcurve . to keep only the important features we define the @xmath13 breakpoint where @xmath14 using a cut - off point it is possible to define a compressed size ( cs ) for a burst : it is the number of bins ( in the wavelet space ) needed to restore the curve at the break .
the cs value is a robust measure quantifying the similarity between the fred kernel and the different channels lightcurves .
our analysis suggest that all the low energy channels # 1 , # 2 and # 3 behaves similarly , while the high energy ( @xmath15 kev ) channel is different ( which is not very surprising , e.g. @xcite ) .
4 . shows the _ ratio _ of the cs s against the total count lightcurves cs for channels 2 and 4 .
this distributions indicate that the pulse - shapes in the long bursts high energy channel are more fred - like than the lower ones - and this is _ independent _ from the actual fred time - scale !
kouveliotou , c. , paciesas , w. s. , fishman , g. j. , meegan , c. a. , & wilson , r. b. 1992 , the compton observatory science workshop , 61 meegan , c. , malozzi , r.s .
, six , f. & connaughton , v. 2001 , current batse gamma - ray burst catalog , http://gammaray.msfc.nasa.gov/batse/grb/catalog norris , j. p. , nemiroff , r. j. , bonnell , j. t. , paciesas , w. s. , kouveliotou , c. , fishman , g. j. , & meegan , c. a. 1994 , american astronomical society meeting , 26 , 1333 norris , j. p. , scargle , j. d. , bonnell , j. t. , & nemiroff , r. j. 1998 , gamma - ray bursts , 4th hunstville symposium , 171 press w.h .
, teukolsky s.a . , vetterling w.t . ,
flannery b.p .
1992 , numerical recipes in fortran , second edition , cambridge university press , cambridge | the gamma - ray burst s lightcurves have been analyzed using a special wavelet transformation . the applied wavelet base is based on a typical fast rise - exponential decay ( fred ) pulse . the shape of the wavelet coefficients total distribution is determined on the observational frequency grid .
our analysis indicates that the pulses in the long bursts high energy channel lightcurves are more fred - like than the lower ones , independently from the actual physical time - scale .
address = laboratory for information technology , etvs university , h-1117 budapest , pzmny p. s. 1./a , hungary address = department of physics , bolyai military university , h-1456 budapest , pob 12 , hungary address = astronomical institute of the charles university , v holeovikch 2 , cz-180 00 prague 8 , czech republic , altaddress = stockholm observatory , albanova , se-106 91 stockholm , sweden address = konkoly observatory , h-1525 budapest , pob 67 , hungary |
we assume here that graphs are finite and loopless .
an _ edge list assignment _ for a graph @xmath1 is a function @xmath5 which maps each edge of @xmath1 a set of colours .
@xmath5-edge colouring _ is a proper edge colouring @xmath10 of @xmath1 for which @xmath11 for each @xmath12 .
let @xmath13 be an edge weighting with positive integers .
we say that @xmath1 is _
@xmath14-edge choosable _ if @xmath1 has an @xmath5-edge colouring for every edge list assignment @xmath5 satisfying @xmath15 for each @xmath16 . for @xmath17 , we say that @xmath1 is _
@xmath18-edge choosable _ if @xmath1 is @xmath14-edge choosable for some @xmath14 satisfying @xmath19 .
if , additionally , @xmath1 has a most @xmath20 edges @xmath8 for which @xmath21 , then we say that @xmath1 is _
@xmath20-nearly @xmath22-edge choosable_. we consider the problem finding a good upper bound on the quantity @xmath23 the notion of _ vertex choosability _ is defined similarly , with reference to vertex colouring .
an extension @xcite of brooks theorem to vertex choosability asserts that every simple connected graph @xmath24 with maximum degree @xmath25 is @xmath25-vertex choosable unless @xmath24 is a complete graph or a cycle .
applying this to the line - graph of a cubic graph , it follows that every loopless cubic graph is @xmath0-edge choosable .
however , a typical cubic graph @xmath1 is @xmath20-nearly @xmath7-edge choosable where @xmath20 is somewhat smaller than the order of @xmath1 .
for example , a cubic graph @xmath1 satisfies @xmath26 ( that is , @xmath1 is @xmath7-edge choosable ) if @xmath1 is either bipartite @xcite , or planar and 2-connected @xcite .
the latter result strengthens the 4-colour theorem . if @xmath1 has a cut - edge , then @xmath27 since @xmath1 is not 3-edge colourable .
one easily constructs cubic graphs @xmath1 having @xmath2 cut - edges for which @xmath28 .
for example , if each connected component of @xmath1 has exactly one cut - edge , then @xmath1 is @xmath14-edge choosable only if each of the @xmath29 leaf - blocks of @xmath1 contains at least one edge @xmath8 for which @xmath30 .
( this is because no leaf block is @xmath7-edge colourable ) . in this paper
we show that planar cubic graphs @xmath1 satisfy @xmath31 .
[ thm : main ] let @xmath1 be a planar cubic graph having @xmath2 cut - edges . then @xmath1 is @xmath14-edge choosable for some function @xmath32 that has average value @xmath7 , and @xmath33 .
let @xmath34 be the edges of a graph @xmath1 , and let @xmath35 be an indeterminate associated with the edge @xmath36 .
the _ edge monomial _ of @xmath1 is the polynomial in @xmath37 $ ] defined by @xmath38 here @xmath39 is the number of vertices incident to both @xmath36 and @xmath40 .
note that @xmath41 is a homogeneous polynomial of degree @xmath42 .
furthermore , @xmath41 is well defined ( up to negation ) regardless of the edge ordering @xmath43 .
each term in the standard expansion of @xmath41 takes the form @xmath44 where the exponent function @xmath45 is a nonnegative integer weighting of the edges of @xmath1 .
we shall write @xmath46 for the function @xmath47 .
the combinatorial nullstellensatz @xcite for edge choosability asserts the following .
[ lem : cn ] let @xmath1 be a loopless graph , and let @xmath48 be a nonzero term in the expansion of @xmath41 .
then @xmath1 is @xmath14-edge choosable , where @xmath49 .
the _ polynomial method _ for proving that @xmath1 is @xmath14-edge choosable typically involves selecting an exponent function @xmath50 satisfying @xmath51 , and showing that @xmath52 . to evaluate @xmath53 , ellingham and goddyn @xcite
provide a combinatorial interpretation of @xmath53 in terms of _ star labellings _ which we describe below .
let @xmath54 be a vertex of degree @xmath55 in @xmath1 .
a _ star labelling at @xmath54 _ is a bijective function @xmath56 from the edges incident with @xmath54 to the integers @xmath57 .
a _ star labelling _ of @xmath1 is a set @xmath58 where each @xmath56 is a star labelling at @xmath54 .
the _ exponent _ of a star labelling @xmath59 is the edge weighting @xmath60 defined by @xmath61 , for @xmath62 .
the _ sign _ , @xmath63 , of a star labelling at @xmath54 is the sign of the permutation of @xmath64 defined by @xmath65 , where @xmath66 are the edges incident with @xmath54 , and @xmath67 .
sign _ of a star labelling of @xmath1 is defined by @xmath68 .
[ lem : combinterp ] for any loopless graph @xmath1 we have @xmath69 where the sum is taken over the set of star labellings of @xmath1 .
the reader should notice that , up to negation , the edge monomial @xmath41 does not depend on the particular ordering @xmath70 of @xmath71 .
a novel feature of this paper is our use of several coefficients of @xmath41 in the polynomial method .
if a set of coefficients @xmath72 has a nonzero sum , then at least one coefficient @xmath73 is not zero .
this gives a multi - term version of lemma [ lem : cn ] .
[ cor : interp ] let @xmath74 be a set of edge weightings for @xmath1 , and let @xmath75 be the set of star labellings @xmath59 of @xmath1 such that the exponent of @xmath59 is a member of @xmath76 . if the integer @xmath77 is not zero , then @xmath1 is @xmath78-edge choosable , for some @xmath79 .
let @xmath80 be the set of star labellings of @xmath1 having exponent @xmath81 . if is not zero , then @xmath82 , for some @xmath83 . by lemma [ lem : combinterp ]
this last sum equals the coefficient of the term @xmath84 in the expansion of @xmath41 , and the result follows from lemma [ lem : cn ] . to prove our main result
, we will construct an appropriate set @xmath76 of edge weightings of a planar cubic graph @xmath1 , and show that the signed sum is positive .
a _ flag _ of a graph @xmath1 is a pair @xmath85 whose members are incident in @xmath1 .
we may write @xmath86 instead of @xmath87 .
it is convenient to regard a star labelling of @xmath1 to be a nonnegative integer function @xmath88 , where @xmath89 is the set of flags of @xmath1 . for @xmath90 ,
a _ thread of order @xmath91 _ is the graph @xmath92 obtained from a path @xmath93 by adding new vertices @xmath94 and new edges @xmath95 ( @xmath96 ) .
see figure [ fig : threadweights ] .
in particular the _ trivial thread _ @xmath97 is the path @xmath98 .
the _ head _ of @xmath92 is the flag @xmath99 , the _ tail _ of @xmath92 is the flag @xmath100 , and the _ feet _ of @xmath92 are the flags @xmath101 ( @xmath96 ) .
a function @xmath102 is called a _ prestar labelling of @xmath92 _ if the restricted function @xmath103 is a star labelling of @xmath104 , for @xmath96 .
sign _ of a prestar labelling @xmath59 is defined to be @xmath105 , and the _ exponent _ of @xmath59 is the edge weighting @xmath50 where @xmath106 for each @xmath107 .
a prestar labelling @xmath59 is _ 1-footed _ if @xmath108 , for @xmath96 .
we say that @xmath59 has _ type @xmath109 _ if @xmath110 and @xmath111 .
we are interested in classifying , according to type , the set of @xmath112-footed prestar labellings of @xmath92 which have a prespecified exponent . for each @xmath113
we define four special edge weightings of @xmath92 , which we denote by @xmath114 , @xmath115 , @xmath116 and @xmath117 .
these are illustrated in figure [ fig : threadweights ] .
the weighting @xmath114 is just the constant function @xmath118 .
the next three weightings are defined only for @xmath119 .
the weighting @xmath115 is obtained from @xmath114 by _ transferring one unit of weight _ from @xmath120 to @xmath121 .
that is , we have @xmath122 , @xmath123 , and @xmath124 for @xmath125 .
the weighting @xmath116 is obtained from @xmath114 by transferring one unit of weight from @xmath126 to @xmath127 .
the weighting @xmath128 is obtained from @xmath114 by transferring one unit of weight from @xmath129 to @xmath126 , and then transferring one unit of weight from @xmath129 to @xmath127 .
as shown in figure [ fig : threadweights ] , each special weighting is associated with one or more @xmath112-footed prestar labellings of @xmath92 . each labelling is denoted by either @xmath130 , @xmath131 or @xmath132 where @xmath109 is its type . , four edge weightings @xmath114 and @xmath133 , and all @xmath112-footed prestar labellings , @xmath130 , @xmath131 and @xmath134 , whose exponents are one of those weights.,width=389 ] [ lem : possiblestars ] there are three prestar labellings @xmath135 , @xmath136 and @xmath137 of @xmath97 with exponent @xmath114 . for @xmath138 , @xmath135 and @xmath136
are the only @xmath112-footed prestar labellings of @xmath92 having exponent @xmath114 .
let @xmath139 and let @xmath140 .
if @xmath141 , then @xmath131 is the unique @xmath112-footed prestar labelling of @xmath92 having exponent @xmath133 . if @xmath142 , then @xmath143 has exactly two 1-footed prestar labellings with exponent @xmath116 , namely @xmath144 and @xmath134 .
we prove only the statement regarding @xmath116 since the arguments are easy and mechanical .
let @xmath59 be a 1-footed prestar labelling of @xmath92 whose exponent equals @xmath116 .
since @xmath59 is @xmath112-footed , we have @xmath145 , and @xmath146 , for @xmath147 . since @xmath148 is a star labelling we have @xmath149 . in case @xmath150
, we have @xmath151 and @xmath152 .
we now apply the facts @xmath153 ( @xmath154 ) , and @xmath155 ( @xmath156 ) to find @xmath157 and @xmath158 for @xmath159 .
thus @xmath160 . in case @xmath161 , we find that @xmath162 , @xmath163 and @xmath164 . if @xmath165 , then we have contradicted the fact @xmath166 is a star labelling , since @xmath167 .
therefore @xmath168 , and @xmath59 is the exceptional prestar labelling @xmath134 .
[ prop : threadsigns ] for each odd integer @xmath119 , the prestar labellings of @xmath92 defined above satisfy @xmath169 . for @xmath168 ,
we have that @xmath170 . for each even integer @xmath165
we have that @xmath171 .
consider the embedding of @xmath92 in the plane shown in figure [ fig : threadweights ] .
the sign of a star labelling at @xmath104 depends only on whether the three labels @xmath172 appear in clockwise or anticlockwise order around @xmath104 .
when @xmath168 , the star labelling at @xmath173 is clockwise under @xmath134 , and is anticlockwise under @xmath144 .
therefore @xmath170 for any embedding of @xmath143 . evidently , every vertex @xmath104 ( @xmath96 ) is anticlockwise under @xmath174 , whereas @xmath173 is the unique anticlockwise vertex under @xmath144 .
therefore @xmath169 provided that @xmath91 is odd .
each vertex @xmath104 is clockwise under @xmath136 and anticlockwise under @xmath135 , so @xmath175 when @xmath91 is even .
we define two variations of an edge - weighted thread . for @xmath139 , the _ closed thread of order @xmath91 _
is the graph @xmath176 obtained from @xmath92 by identifying the vertices @xmath177 and @xmath178 .
we define the edge weightings @xmath116 , @xmath114 , and the prestar labellings @xmath144 , @xmath135 and @xmath136 exactly as they were defined for @xmath92 .
the reader will easily verify the following lemma .
[ lem : uniq - circ ] let @xmath138 .
then @xmath135 and @xmath136 are the only 1-footed prestar labellings of @xmath176 having exponent @xmath114 .
furthermore , @xmath144 is the unique 1-footed prestar labelling of @xmath176 having exponent @xmath116 , for which the head and tail ( that is , the two flags incident with @xmath177 ) receive different labels .
an _ injured thread of order @xmath91 _ is any graph @xmath179 that is obtained from @xmath92 by deleting any one of its @xmath91 `` feet '' @xmath94 , @xmath96 .
the edge weighting @xmath180 of @xmath179 is the restriction of @xmath115 to the edge set of @xmath179 .
@xmath112-footed prestar labelling _ of @xmath179 is the restriction of a @xmath112-footed prestar labelling of @xmath92 to the flags in @xmath179 . in particular , we define @xmath181 . to ease notation , we shall write @xmath115 instead of @xmath180 , and write @xmath182 instead of @xmath183 , where no confusion results .
[ lem : uniq - deleted ] for @xmath138 , the prestar labelling @xmath182 is the unique 1-footed prestar labelling of @xmath179 whose exponent equals @xmath115 .
in this section , we define a set of edge weightings @xmath76 of a planar cubic graph , to which we will apply corollary [ cor : interp ] .
let @xmath1 be a connected loopless cubic graph and let @xmath184 be the set of cut - edges in @xmath1 .
a _ block _ of @xmath1 is any connected component of @xmath185 ( this differs from the standard definition of `` block '' ) .
each block , @xmath24 , is either a _ vertex block _ , a _ cycle block _ or a _ proper block _ , depending on whether @xmath24 is a single vertex , a cycle or a subdivision of @xmath186-connected cubic graph .
the _ block tree _ of @xmath1 is the tree obtained by contracting each block @xmath24 to a single vertex , which we also denote by @xmath24 where no confusion results .
since @xmath1 is finite , at least one block of @xmath1 is a proper block .
we designate one proper block to be the @xmath187 @xmath188 of @xmath1 .
every other block of @xmath1 is called a _ nonroot block _ of @xmath1 .
we define @xmath189 to be the set of edges in @xmath184 which have exactly one end in @xmath188 .
every nonroot block @xmath24 is incident to a unique cut - edge , denoted by @xmath190 , which lies on the path from @xmath24 to @xmath191 in the block tree of @xmath1 . for nonroot blocks
@xmath24 , we define @xmath192 to be the set of edges in @xmath193 which have an endpoint in @xmath24 . for any block @xmath24 of @xmath1 ,
let @xmath194 be the subgraph of @xmath1 obtained from @xmath24 by adding all the edges in @xmath192 and their endpoints .
each subgraph @xmath194 is called an _ extended block _ of @xmath1 . the extended blocks of @xmath1
depend on the choice of @xmath188 .
the edge sets of the extended blocks of @xmath1 form a partition of @xmath71 .
we further refine the extended blocks into pieces that are each isomorphic to one of the threads , @xmath92 , @xmath195 or @xmath196 , defined in section [ sec : threads ] .
each extended vertex block is a path of length @xmath186 , which we regard to be copy of the injured thread @xmath197 .
we define the family @xmath198 each extended cycle block is isomorphic to a closed thread @xmath176 , for some @xmath138 .
we group these into two families .
@xmath199 the reader should notice that if @xmath200 , then the length of the cycle @xmath24 has opposite parity to @xmath91 .
each extended proper block @xmath194 of @xmath1 decomposes into copies of threads @xmath92 and injured threads @xmath196 as follows . by suppressing every vertex of degree @xmath186 in @xmath24 we obtain a @xmath186-connected cubic graph homeomorphic to @xmath24 , which is denoted @xmath201 and called the _ derived graph _ of @xmath24 .
each edge @xmath202 corresponds to a maximal induced path @xmath203 of positive length in @xmath24 . by adding to @xmath203 those edges in @xmath192 ( and their endpoints ) which are incident to @xmath203
, we obtain a subgraph @xmath204 which is isomorphic to either a thread @xmath92 , @xmath113 , or injured thread @xmath196 , @xmath138 .
the edge sets of the subraphs in @xmath205 form a partition of @xmath206 .
summarizing , we have decomposed @xmath1 into a family @xmath207 of copies of threads , injured threads and closed threads where @xmath208 the members of @xmath209 are called _ general threads of @xmath1 _ , and @xmath210 is the _ derived graph _ of @xmath1 .
an edge @xmath211 is a _ base edge _ of @xmath210 if @xmath212 is isomorphic to an injured thread .
thus each connected component of @xmath210 other than @xmath213 contains exactly one base edge .
each @xmath214 has zero or more well defined _ feet _ , but there are two ways to select which end is the _ head _ of @xmath215 .
let @xmath1 be a connected planar cubic graph where a base block @xmath188 has been selected .
let the @xmath209 and @xmath210 be the general thread decomposition and reduced graph as defined above . to describe
a weighting of @xmath1 it suffices to specify , for each @xmath216 , which end of @xmath215 is the head , and which of the weightings described in section [ sec : threads ] to assign to @xmath215
. this specification will make reference to a particular perfect matching in the reduced cubic graph @xmath210 .
if @xmath217 is a perfect matching in @xmath210 , then the edge set @xmath218 is a _ @xmath186-factor _ of @xmath210 .
we say that @xmath219 is _ bipartite _ if every cycle of @xmath220 has even length .
let @xmath221 be a perfect matching in @xmath210 satisfying the following properties . 1 .
every base edge of @xmath210 is an edge in @xmath222 , [ m:1 ] 2 .
the @xmath186-factor @xmath223 is bipartite , 3 .
subject to conditions ( 1 ) and ( 2 ) , @xmath222 contains the maximum possible number of edges @xmath224 for which the general thread @xmath212 is nontrivial ( that is , @xmath225 ) .
the matching @xmath222 exists because every component of @xmath210 has a proper @xmath7-edge colouring ( by the four colour theorem ) , and has at most one base edge .
we partition the set of threads @xmath226 into five classes @xmath227 where @xmath228 we have defined the following partition of the general threads of @xmath1 into eight classes .
@xmath229 by the choice of @xmath222 , every injured thread in @xmath209 belongs to @xmath230 .
let @xmath231 be any fixed cyclic orientation of the @xmath186-factor @xmath232 .
for each general thread in we arbitrarily designate one of two possible flags to be its head , subject to the following condition .
@xmath233 we now refer to the thread weightings defined in section [ sec : threads ] . for every subset @xmath234 , we define @xmath235 to be the edge weighting which restricts to every general thread @xmath216 , as follows .
@xmath236 finally , we define the following set of edge weightings of @xmath1 , @xmath237
let @xmath1 be a planar cubic graph . we designate a proper block of @xmath1 to be the root block of @xmath1 .
we define the extended blocks , and @xmath210 , @xmath222 , @xmath231 and @xmath238 as in section [ sec : weightset ] .
let @xmath239 be the set of star labellings of @xmath1 whose exponent is @xmath240 , and let @xmath241 a _ base flag _ of @xmath1 is any flag that is a foot of some general thread in @xmath1 . thus every base flag takes the form @xmath242 where @xmath243 is the unique vertex of degree 2 in some nonroot extended block @xmath194 , and @xmath190 is the unique cut - edge of @xmath1 which is incident to @xmath243 , and is not an edge of @xmath194 .
[ prop:1-footed ] for every star labelling @xmath244 we have @xmath245 , for every base flag @xmath246 of @xmath1 .
let @xmath247 be the exponent of @xmath59 .
let @xmath246 be a base flag in @xmath1 , and let @xmath248 be the set of blocks @xmath249 of @xmath1 for which @xmath190 lies on the unique path from @xmath249 to the root block of @xmath1 in the block tree of @xmath1 .
let @xmath250 .
every vertex in the subgraph @xmath5 has degree @xmath7 except for @xmath243 , which has degree @xmath186 . for every extended block @xmath251 of @xmath1 , the average value of @xmath252 among the edges @xmath253 equals @xmath186
this is because each of the weightings @xmath254 has average value @xmath186 in the definition of @xmath240 .
since @xmath255 is a partition of @xmath256 , the average value of @xmath257 among the edges of @xmath5 equals @xmath186 .
therefore the average value of @xmath258 among the flags in @xmath259 equals @xmath112 . on the other hand ,
each vertex of @xmath5 is incident to three flags in @xmath260 and contributes @xmath261 to the value of @xmath262 .
thus the average value of @xmath258 among the flags in @xmath260 equals @xmath112 , and @xmath245 .
[ cor : restrictedsl ] let @xmath234 .
then every star labelling @xmath263 , satisfies the following . 1 . for every @xmath264 we have @xmath265 .
[ choice:11 ] 2 . for every @xmath266
, we have @xmath267 .
[ choice:20 ] 3 . for every @xmath268
, we have @xmath269 . [ choice:02 ] 4 . for every @xmath270
, we have @xmath271 . [ choice:02circ ] 5 . for every @xmath272 , we have @xmath273 . [ choice:2 ]
6 . for every @xmath274
, we have @xmath275 . [ choice:0 ] for any general thread @xmath216 , the restriction @xmath276 is 1-footed , by proposition [ prop:1-footed ] . now all the statements except follow immediately from the definition of @xmath240 and the three lemmas in section [ sec : threads ] . for the statement
, we observe that the head and tail of a circular thread in must receive distinct labels in any star labelling of @xmath1 , and the claim follows from lemma [ lem : uniq - circ ] . for every star labelling @xmath59 of @xmath1
, we define a a corresponding star labelling @xmath277 of @xmath210 called the _ derived _ star labelling .
informally , @xmath278 is the restriction of @xmath59 to the heads and the tails of the general threads of @xmath1 .
more precisely , for each @xmath211 , let @xmath279 be the corresponding general thread of @xmath1 .
let @xmath280 and @xmath54 be the endpoints of @xmath224 that correspond to the head and tail of @xmath215 , respectively .
the restriction @xmath276 corresponds a 1-footed prestar labelling of a thread or injured thread having type @xmath109 , for some @xmath281 .
we define @xmath282 and @xmath283 .
let @xmath244 . by the definition of @xmath76 ,
the exponent of @xmath278 is the constant function @xmath284 , @xmath285 .
let @xmath286 be the set of edges @xmath287 such that @xmath288 .
let @xmath289 and let @xmath290 be the orientation of @xmath291 where , for @xmath292 we have @xmath293 and @xmath294 .
then @xmath286 is a perfect matching of @xmath210 , and @xmath290 is an oriented @xmath186-factor of @xmath210 .
the correspondence between the star labellings of @xmath210 with exponent @xmath295 and the pairs @xmath296 where @xmath297 is an oriented @xmath186-factor of @xmath210 is bijective .
the following will help us later deal with the exceptional star labelling @xmath134 that arises in part of corollary [ cor : restrictedsl ] .
[ lem : pi11oddcycle ] let @xmath234 and let @xmath263 .
let @xmath291 be the @xmath186-factor of @xmath210 as defined above . if @xmath1 has a general thread @xmath216 for which @xmath298 , then some connected component of @xmath291 is an odd cycle in @xmath210 .
let @xmath215 be as in the statement . by corollary [ cor : restrictedsl ]
we necessarily have @xmath299 let @xmath300 be the set of general threads in @xmath301 which are nontrivial and receive a type @xmath302 prestar labelling under @xmath59 .
by part of corollary [ cor : restrictedsl ] , we have @xmath303 . by the hypothesis
, we also have @xmath304 so @xmath305 . if the @xmath186-factor @xmath291 of @xmath210 were bipartite , then the perfect matching @xmath286 would contradict our choice of @xmath222
. therefore some component of @xmath291 is an odd cycle of @xmath210 .
let @xmath306 be the perfect matching in @xmath210 and the orientation of the complementary 2-factor used in the definition of the set of weightings @xmath76 .
let @xmath307 be the unique star labelling of @xmath1 satisfying * @xmath308 , * @xmath309 , * every closed thread @xmath310 receives the labelling @xmath136 , as in of corollary [ cor : restrictedsl ] .
the star labelling @xmath307 is called the _ reference star labelling _ of @xmath1 . accordingly
, the reduced star labelling @xmath311 is called the _ reference star labelling _ of @xmath210 .
these star labellings will be used for sign computations .
it convenient to assume that @xmath312 .
let @xmath313 be the set of star labellings @xmath314 for which @xmath291 is a bipartate @xmath186-factor of @xmath210 .
in particular , @xmath315 because @xmath316 is bipartite .
the following was proved by ellingham and goddyn @xcite .
[ eg ] let @xmath278 be a star labelling of a planar cubic graph @xmath210 with exponent @xmath317 .
if @xmath291 is bipartite , then @xmath318 .
[ cor : sgn ] let @xmath319 .
then @xmath320 , where @xmath321 is the number of threads @xmath322 for which @xmath323 .
we have @xmath324 .
the set of prestar labellings of @xmath216 whose exponents are determined by a weighting in @xmath76 are restricted according to statements ( 1 ) to ( 6 ) of corollary [ cor : restrictedsl ] . by proposition [ prop : threadsigns ] , both prestar labellings listed in statement have the same sign , whereas the two labellings in statement have opposite sign . in statements , and , the prestar labelling is fixed , and in the sign is @xmath112 since the threads there are trivial . by lemma [ eg ] , @xmath318 for @xmath319 .
the result follows from the facts that @xmath325 for @xmath326 , and that @xmath327 .
let @xmath328 .
we now define a particular function @xmath329 which maps each member of @xmath330 to another star labelling of @xmath1 .
we fix an arbitrary total ordering of the set of odd cycles in @xmath210 . for @xmath331 ,
let @xmath332 be the first odd cycle which is a component of @xmath333 $ ] .
every general thread @xmath334 has type @xmath335 or type @xmath336 , so we have @xmath337 , and one of the cases , , or of corollary [ cor : restrictedsl ] applies to @xmath215 .
( we used the fact @xmath134 has type @xmath302 . )
let @xmath338 be the star labelling in @xmath339 obtained from @xmath59 as follows .
for every @xmath340 we do the following .
if @xmath279 is trivial , then we interchange the labels @xmath341 and @xmath186 on its two flags . otherwise , we relabel the flags of @xmath215 in a way that interchanges either the prestar labellings @xmath174 and @xmath144 ( if @xmath342 ) , or the prestar labellings @xmath135 and @xmath343 ( if @xmath344 ) .
more precisely , for @xmath345 , if @xmath346 , then @xmath347 , and if @xmath348 , then @xmath349 .
[ prop : involution ] the map @xmath14 is a fixed - point free involution @xmath350 which satisfies @xmath351 .
let @xmath331 and let @xmath352 be the exponent of @xmath59 .
then the exponent of @xmath338 is the weighting @xmath353 where @xmath354 is the symmetric difference of @xmath355 and @xmath356 .
therefore we have @xmath357 .
clearly @xmath358 and @xmath359 so @xmath14 is a fixed - point free involution on @xmath339 . for @xmath360 ,
the star labellings @xmath56 and @xmath361 differ by the transposition @xmath362 , whereas @xmath363 for every @xmath364 .
since @xmath332 has odd length , the derived star labellings therefore satisfy @xmath365 .
the result now follows from corollaries [ cor : sgn ] and [ prop : threadsigns ] .
we note that the oriented @xmath186-factor @xmath366 is obtained from @xmath290 by reversing all the arcs in the odd cycle @xmath332 .
let the set of edge weights @xmath76 be defined as in section [ sec : weightset ] , and let @xmath367 be the star labellings of @xmath1 with exponent in @xmath76 , as defined in section [ sec : starlabellings ] .
let @xmath319 .
then @xmath291 is a bipartite @xmath186-factor of @xmath210 . applying lemma [ lem : pi11oddcycle ] , we conclude that no general thread @xmath215 satisfies @xmath323 .
it follows from corollary [ cor : sgn ] that @xmath368 for every @xmath319 .
we have that @xmath369 , since @xmath370 contains the reference star labelling @xmath307 .
therefore @xmath371 .
we have by proposition [ prop : involution ] that @xmath372 .
thus we have shown that @xmath373 .
applying corollary [ cor : interp ] we have that @xmath1 is @xmath374-edge choosable for some @xmath375 .
we are interested in the upper bound @xmath376 .
suppose @xmath1 has @xmath2 cut - edges .
for any general thread @xmath216 , let @xmath377 let @xmath281 .
define @xmath378 , let @xmath379 , and let @xmath380 . since every thread in @xmath381 has positive length we have @xmath382 .
let @xmath8 be a cut - edge of @xmath1 .
there are exactly two general threads @xmath383 such that @xmath8 joins a vertex of degree @xmath384 in @xmath215 to a vertex of degree @xmath385 in @xmath386
. therefore each cut - edge @xmath8 contributes exactly twice to the quantity @xmath387 , so @xmath388 furthermore , at least one of the two contributions of @xmath8 goes toward @xmath389 , because the thread in @xmath390 that lies farther from the root block @xmath188 is always a member of @xmath391 . therefore @xmath392 . by examining figure [ fig : threadweights ]
we find that @xmath393 . to obtain the claimed upper bound of @xmath4
, we must argue more carefully .
let @xmath394 be the edge weighting of @xmath1 obtained from @xmath395 by interchanging the head and tail of every thread @xmath396 ( see in section [ sec : weightset ] ) , and then swapping the roles of `` @xmath355 '' and `` @xmath397 '' in the definition of @xmath240 .
more precisely , the weightings @xmath50 and @xmath394 are identical , except that for every thread @xmath396 , exactly one of the restricted weightings in @xmath398 coincides with @xmath116 , and the other coincides with @xmath117 after exchanging the head and tail of @xmath215 .
then the coefficients of @xmath399 and @xmath400 in @xmath41 are equal in absolute value .
this is because there is a natural bijection from the star labellings of @xmath1 with exponent @xmath50 to those with exponent @xmath394 . in the bijection ,
each star labelling @xmath59 with exponent @xmath50 maps to the unique star labelling @xmath401 with exponent @xmath394 which is identical on all flags outside of any thread in @xmath402 , and for which the reduced star labellings of @xmath210 satisfy @xmath403 .
we now compute @xmath404 .
the analogue of is that @xmath405 .
we sum these two equations .
@xmath406 we apply the inequality just before .
@xmath407 so @xmath408 . that is to say ,
that one of the two functions , @xmath409 or @xmath410 , satisfies the statement of theorem [ thm : main ] .
our use of the weightings @xmath114 in the definition of @xmath240 was included as a mild attempt to minimize the number of edges of @xmath1 receiving weight @xmath7 . in particular ,
extended cycle blocks of even length , ( open ) threads of even order , and edges in trivial threads do not require lists of length @xmath0 . by inspecting the above inequalities , it is apparent that the upper bound @xmath411 can be improved if many threads of @xmath1 have length @xmath412 , or if @xmath210 has a proper @xmath7-edge colouring in which most of the edges coming from nontrivial threads have the same colour .
it would be very interesting to improve the bounds @xmath413 , where @xmath414 .
similar questions can be asked regarding regular planar graphs of higher degree .
n. alon , restricted colorings of graphs , _ in _ `` surveys in combinatorics '' , proc .
@xmath415 british combinatorial conference , london mathematical society lecture notes series 187 , edited by k. walker , cambridge university press , 1993 , 1 - 33 . | it is known that every loopless cubic graph is @xmath0-edge choosable .
we prove the following strengthened result .
let @xmath1 be a planar cubic graph having @xmath2 cut - edges .
there exists a set @xmath3 of at most @xmath4 edges of @xmath1 with the following property . for any function @xmath5 which assigns to each edge of @xmath3 a set of @xmath0 colours and which assigns to each edge in @xmath6 a set of @xmath7 colours , the graph @xmath1 has a proper edge colouring where the colour of each edge @xmath8 belongs to @xmath9 . |
the connes kreimer hopf algebra and the butcher group .
the hopf algebra @xmath5 of rooted trees is now a well - established object in mathematics , thanks in particular to the seminal works of connes and kreimer .
kreimer @xcite discovered that @xmath5 controls the combinatorics of renormalisation in perturbative quantum field theory , and his collaboration with connes ( e.g. @xcite , @xcite , to cite a few ) uncovered deep connections with noncommutative geometry , number theory , lie theory , and algebraic combinatorics , stimulating a lot of further activity by many mathematicians and physicists . the connes
kreimer hopf algebra has now been characterised by several different universal properties @xcite , @xcite , @xcite .
the group of characters of @xmath5 , now called the butcher group , was in fact studied by butcher @xcite some 30 years earlier , in relation with order conditions for runge kutta methods in numerical integration .
the link back to this work from the connes
kreimer hopf algebra was provided by brouder @xcite .
categorification of the connes kreimer hopf algebra .
this article investigates a categorification of the bialgebra @xmath5 , i.e. a lift from the level of algebra to the level of sets . to be precise
, the antipode is _ not _ categorified .
indeed , the antipode is a feature depending on the additive inverses of the ground ring .
but in fact , the most natural ` ring ' of definition of @xmath5 is the semiring @xmath6 of natural numbers , and lacking additive inverses there is no antipode to be had here .
the semiring @xmath6 appears as the burnside semiring of the distributive category @xmath7 of finite sets , i.e. is the set of isomorphism classes of finite sets , with addition and multiplication inherited from the categorical sum and product .
exhibiting a distributive category whose burnside semiring is @xmath5 , and describing at this level the comultiplication , is what we mean by categorification , following a popular terminology which goes back to crane and yetter @xcite ; the specific process employed can more precisely be called objectification , cf .
lawvere , schanuel , and their collaborators .
a recommended introduction to categorification is the beautiful paper @xcite of baez and dolan .
while the combinatorial nature of @xmath5 makes it clear that this categorification should be possible , a considerable amount of categorical algebra is needed in order to make all the algebraic structure explicit at the objective level .
the categorification of @xmath5 will be the distributive category @xmath8 $ ] of polynomial functors on the category @xmath9 of @xmath10-indexed finite sets , where @xmath10 is the set of trees .
for technical reasons we mostly work with @xmath0-trees for a finitary polynomial endofunctor @xmath0 , and we start out by explaining the differences
. the comultiplication will be a comomonoidal structure on @xmath8 $ ] , relative to a certain tensor product of distributive categories .
it is conceptually much easier to take the dual viewpoint
. the category @xmath9 will be the categorification of the butcher group ( or rather the butcher monoid ) , and the task is to describe the monoidal structure on @xmath9 dual to the comultiplication .
this monoidal structure , a functor @xmath11 , is itself a polynomial functor . the comultiplication is now given by precomposition with @xmath12 .
polynomial functors .
the notion of polynomial functor is central to this work .
the theory of polynomial functors has roots in topology , representation theory , combinatorics , logic and computer science , but the task of unifying these developments has only recently begun @xcite .
an important feature of polynomial functors is that they can be manipulated in terms of a few representing sets , just as polynomial functions can be manipulated in terms of their coefficients and exponents . in the present work ,
polynomial functors enter at two levels : firstly , and most importantly , the notion of polynomial functor categorifies the notion of polynomial function ; second , the trees that index the involved variables are operadic trees , and they are themselves defined in terms of polynomial endofunctors . for the first aspect we need to develop some basic theory , which constitutes section [ sec : alg ] ; for the second , the theory needed is already available from @xcite and @xcite .
free monads , @xmath0-trees , and beyond .
historically , one starting point for the general project of categorification is the quest in combinatorics for bijective proofs : it is well appreciated that a bijection between sets represents better understanding than a mere equation between numbers .
one insight into the hopf algebra of rooted trees which results from its categorification and the polynomial viewpoint is the relation with free monads : it is shown that the set maps occurring in the polynomial representation of the new monoidal structure are the same maps as occur in the polynomial representation of the free monad construction .
unfortunately , this is not completely true for abstract trees .
it is true for @xmath0-trees , and we develop the theory in this setting the free monad in question is then the free monad on @xmath0 .
the notion of @xmath0-tree ( cf .
[ ptree ] below ) covers many notions of structured and decorated trees , such as binary and planar trees , but abstract trees themselves are not an example : abstract trees should be @xmath0-trees for the terminal polynomial endofunctor , but the category of polynomial endofunctors over sets ( see [ cart ] ) does not have a terminal object !
nevertheless , with a little care , the constructions made for @xmath0-trees work also for abstract trees , only the relation with free monads is then less direct .
this is explained in section [ sec : beyond ] , where it is also explained how these issues disappear when upgrading the whole theory from sets to groupoids : in this fancier setting , the terminal polynomial endofunctor does exist , and abstract trees become a particular case of the notion of @xmath0-tree .
outline of the paper .
in section [ sec : bialg ] we introduce the bialgebra of operadic trees , and explain its relation with the usual connes kreimer hopf algebra .
section [ sec : poly ] collects the notions and results needed from the theory of polynomial functors , notably the explicit formula for substitution of polynomials . in section [ sec : alg ] we set up a framework for dealing with categories of polynomial functors and categories of indexed finite sets as ` polynomial rings ' and ` affine space ' , respectively .
this section contains many observations that seem not to have been made before , regarding polynomial functors as categorification of polynomial functions . in section [ sec : trees ] we introduce trees and @xmath0-trees and review in detail the construction of the free monad on a polynomial endofunctor @xmath0 . finally in section [ sec : monoidal ] we establish the monoidal structure on @xmath9 , and relate it to the bialgebra of trees .
section [ sec : beyond ] contains some remarks about the difference between @xmath0-trees and abstract trees , and hints at a groupoid version of all the constructions as a more comprehensive framework .
related work .
this work was presented at the 2010 international category theory conference in genova . at the same conference ,
matas menni spoke about his work with lawvere @xcite on categorification of incidence algebras of mbius categories . while it is known that the connes
kreimer hopf algebra can be constructed as the incidence algebra of a suitable family of posets , it seems unlikely that it can be given as the incidence algebra of a single mbius category , so at present the lawvere
menni approach does not apply to categorify the connes
kreimer hopf algebra .
work is under way to develop a higher - categorical notion of mbius categories in order to unify the two approaches .
acknowledgments . the author has benefited greatly from many conversations with kurusch ebrahimi - fard , and
also thanks the anonymous referees for many pertinent remarks which led to improved exposition .
partial support from research grants mtm2009 - 10359 and mtm2010 - 20692 of spain and 2009sgr-1092 of catalonia is gratefully acknowledged .
the standard connes kreimer hopf algebra of rooted trees concerns combinatorial trees , whereas in this paper we prefer to work with operadic trees .
this section explains the differences .
combinatorial trees .
the trees usually employed , which here we call combinatorial trees , are often defined as finite connected graphs without loops or cycles , and with a designated root vertex . if a connected subgraph of a rooted tree @xmath13 does not contain the root of @xmath13 , then instead it has a vertex nearest the root , which is then defined to be its root . the connes
kreimer bialgebra of rooted trees .
the bialgebra of rooted trees of connes and kreimer @xcite is the polynomial @xmath14-algebra @xmath5 on the set of isomorphism classes of combinatorial trees . here
@xmath14 can be any commutative @xmath6-algebra .
the comultiplication is given on generators by @xmath15 where the sum is over all admissible cuts of @xmath13 ; the left - hand factor @xmath16 is the forest ( interpreted as a monomial ) found above the cut , and @xmath17 is the subtree found below the cut ( or the empty forest , in case the cut is below the root ) . `
admissible cut ' means upper - set in the poset underlying the tree ( oriented from leaves ( inputs ) to root ( output ) ) i.e. either a subtree containing the root , or the empty set .
@xmath5 is a connected bialgebra : the grading is by the number of nodes , and @xmath18 is spanned by the unit .
therefore , by general principles ( see for example @xcite ) , it acquires an antipode and becomes a hopf algebra provided @xmath14 has additive inverses , i.e. is a @xmath3-algebra .
( in any case , the antipode exists after base change to @xmath3 . ) operadic trees .
we shall need trees with slightly more expressive power , by allowing loose ends ( leaves ) : these are _ operadic trees _ , also called finite rooted trees with boundary a formal definition is given in [ polytree - def ] .
for the moment , the following drawings should suffice to exemplify operadic trees as usual the planar aspect inherent in a drawing should be disregarded : 0.5 ( -50 0 ) ( 0 0 ) ( 0 30 ) ( 0 0 ) ( 0 0 ) ( 0 18 ) ( 0 0 ) f:0 r:2 ( 50 0 ) ( 0 0 ) ( 0 36 ) ( 0 18 ) ( 0 0 ) f:0 r:2 ( 105 0 ) ( 0 0 ) ( 0 15 ) ( 0 0 ) f:0 r:2 ( -5 33 ) ( 0 0 ) f:0 r:2 ( 0 15 ) ( -12 28 ) ( 0 0 ) f:0 r:2 ( 0 15 ) ( 4 43 ) ( 0 15 ) ( 12 40 ) ( 170 0 ) ( 0 0 ) ( 0 18 ) ( 0 0 ) f:0 r:2 ( -6 32 ) ( 0 0 ) f:0 r:2 ( -12 57 ) ( 0 18 ) ( 4 40 ) ( 0 0 ) f:0 r:2 ( 20 50 ) ( 0 0 ) f:0 r:2 ( 15 65 ) ( 20 50 ) ( 25 65 ) ( 4 40 ) ( 9 54 ) ( 0 0 ) f:0 r:2 ( 4 40 ) ( -4 61 ) note that certain edges ( the _ leaves _ ) do not start in a node and that one edge ( the obligatory _ root edge _ ) does not end in a node . a node without incoming edges is not the same thing as a leaf ; it is a nullary operation ( i.e. a constant ) in the sense of operads .
in operad theory , the nodes represent operations , and trees are formal combinations of operations .
the small incoming edges drawn at every node serve to keep track of the arities of the operations . furthermore , for coloured operads , the operations have type constraints on their inputs , encoded as attributes of the edges .
operadic trees in qft .
the use in quantum field theory of more refined notions of trees , and operadic trees in particular , has been hinted at by kreimer in several papers , most concretely with bergbauer @xcite , where trees with loose edges are used to analyse combinatorial dyson schwinger equations .
in fact , the role of trees in the connes kreimer bialgebra is to encode nestings of 1-particle irreducible feynman graphs @xcite , and one can argue @xcite that they are naturally operadic : each tree naturally comes equipped with decorations by primitive 1pi graphs on nodes and interaction labels on edges . to fully encode the compatibility conditions involved in these decorations , and to allow to recover the graph from the decorated trees , it is necessary to keep track of the arities of the nodes , so that even vacant input slots are represented
; this leads naturally to operadic trees .
a more thorough analysis of the relationship between graphs and trees is given elsewhere @xcite .
kreimer @xcite stresses the general utility of trees as expression of nestings of structures , not only of feynman graphs . in the following picture
we see first a nesting of subsets , then a combinatorial - tree expression of the nesting , and finally an operadic - tree expression of the same nesting , in which the leaves correspond to the elements of the nested sets : 0.5 ( 0 30 ) r:30 ( -15 15 ) r:7 ( 8 32 ) r:19 ( 17 38 ) r:5 ( 7 24 ) r:9 ( -15 15 ) ( 0 0 ) f:0 r:2 ( 4 25 ) ( 0 0 ) f:0 r:2 ( 10 23 ) ( 0 0 ) f:0 r:2 ( 4 42 ) ( 0 0 ) f:0 r:2 ( -13 48 ) ( 0 0 ) f:0 r:2 0.8 ( 0 15 ) 1.5 ( 0 0 ) f:0 r:3 ( -10 20 ) f:0 r:3 ( 0 0 ) ( 10 20 ) f:0 r:3 ( 0 40 ) f:0 r:3 ( 10 20 ) ( 20 40 ) f:0 r:3 ( 0 0)(*ex * ey ) ( * ex * ey ) ( 0 0 ) 0.8 1.5 ( 0 0 ) f:0 r:3 ( -10 20 ) f:0 r:3 ( 0 0 ) ( 10 20 ) f:0 r:3 ( 0 40 ) f:0 r:3 ( 10 20 ) ( 20 40 ) f:0 r:3 ( 0 0)(*ex * ey ) ( * ex * ey ) 0.5 ( 0 40 ) ( -5 65 ) ( 0 40 ) ( 5 65 ) ( 0 0 ) ( 0 -15 ) ( 0 0 ) ( -40 45 ) ( -10 20 ) ( -22 60 ) ( 10 20 ) ( 15 64 ) ( 0 0)(*ex * ey ) ( * ex * ey ) one feature of operadic trees is that they admit colimit descriptions of basic operations : most importantly , grafting can be expressed as a pushout , and every tree is the colimit of its elementary subtrees , i.e. trees without inner edges @xcite .
this makes them well suited for constructions ( rather than just decomposition ) .
another advantage is that symmetries of the original nested structure ( for example a nesting of feynman graphs ) are better captured by operadic trees than by combinatorial trees , as the above picture also illustrates : the combinatorial tree has a symmetry which does not reflect a symmetry in the nesting , and fails to detect the inner symmetries of the nesting .
( the symmetry issues play a crucial role in the treatment of green functions , where the operadic viewpoint seems important @xcite . )
the bialgebra of operadic trees.[comultmy ] a _ cut _ of an operadic tree is defined to be a subtree containing the root note that the arrows in the category of operadic trees ( section [ sec : trees ] ) are arity preserving ( [ sub ] ) , meaning that if a node is in the subtree , then so are all the incident edges of that node .
if @xmath19 is a subtree containing the root , then each leaf @xmath20 of @xmath21 determines an _ ideal subtree _ of @xmath13 ( [ sub ] ) , namely consisting of @xmath20 ( which becomes the new root ) and all the descendant edges and nodes .
this is still true when @xmath20 is also a leaf of @xmath13 : in this case , the ideal tree is the trivial tree consisting solely of @xmath20 .
figuratively , this means that cuts can go through the leaves but are not allowed to go _ above _ the leaves .
note also that the root edge is a subtree ; the ideal tree of the root edge is of course the tree itself .
this is the analogue of the cut - below - the - root in the combinatorial case . for a cut @xmath19 ,
define @xmath16 to be the forest consisting of all the ideal trees generated by the leaves of @xmath21 .
let @xmath22 be the polynomial @xmath14-algebra on the set of isomorphism classes of operadic trees .
with comultiplication defined on the generators by @xmath23 as for combinatorial trees , @xmath22 becomes a graded bialgebra .
it is not connected : @xmath24 is spanned by powers of the trivial tree @xmath25 ( including the empty power , which is the algebra unit @xmath26 ) .
these are all group - like , so one could obtain a connected bialgebra by imposing the equation @xmath27 . note that the comultiplication for operadic trees is linear in the right - hand factor , unlike the comultiplication for combinatorial trees , where @xmath28 always contains a factor @xmath29 corresponding to the empty - cut - below - the - root . as an example , here are the 11 possible cuts of the tree in the previous picture : note that the first and the last term of the comultiplication of this tree @xmath13 are @xmath30 not @xmath31 . from operadic trees to combinatorial trees .
the _ core _ of a non - trivial operadic tree @xmath13 is the combinatorial tree @xmath32 obtained by pruning off all leaves as well as the root edge .
taking core is functorial in root - preserving inclusions .
for any non - trivial operadic tree @xmath13 , there is a natural bijection between the set of non - trivial subtrees of @xmath13 and the set of combinatorial subtrees of @xmath32 .
this bijection sends @xmath33 to @xmath34 . by extending the assignment by defining the core of the trivial tree to be the empty ( combinatorial ) forest ,
it is clear that the operation is compatible with comultiplication , as illustrated in the previous picture , where the core is highlighted with fat lines .
in conclusion we have : proposition . _
taking core defines a surjective homomorphism of graded bialgebras @xmath35 _ the bialgebra of @xmath0-trees .
the definition makes sense equally well for @xmath0-trees ( [ ptree ] ) .
if there are more than one edge colour involved , there is a trivial tree for each edge colour , and the monomials in these trees ( including the trivial monomial @xmath26 ) span @xmath24 .
clearly they are all group - like .
in this section , we recall the basic notions of polynomial functors , referring to gambino
kock @xcite for all details .
slices , pullback , and adjoints .
we write @xmath36 and @xmath37 for the disjoint union of sets ( i.e. the categorical coproduct ) .
the equality sign denotes canonical isomorphism of sets .
let @xmath38 be a set .
recall ( e.g. from @xcite , 9.7 ) the _ slice category _ @xmath39 : its objects are maps @xmath40 , and its arrows are commutative triangles x & & & & x + & & & & + & & b . & & for a set map @xmath41 , we denote the fibre over @xmath42 by @xmath43 .
we can then write @xmath44 as the union of the fibres , @xmath45 , and interpret @xmath46 as a @xmath38-indexed family of sets .
in fact , there is a canonical equivalence of categories @xmath47 & \longmapsto & [ b\mapsto x_b ] .
\end{aligned}\ ] ] here @xmath48 denotes the category of functors @xmath49 ( considering the set @xmath38 as a discrete category ) . for either interpretation of @xmath38-indexed family of sets
, we shall use the notation @xmath50 . given a map @xmath51
, we have the _ lowershriek _ functor @xmath52 & \longmapsto & [ x\to a \to b ] .
\end{aligned}\ ] ] using decomposition into fibres , the functor has the following description : @xmath53 i.e. sum along the fibres .
the functor @xmath54 has a right adjoint , denoted by _
upperstar _ , given by pullback : @xmath55[0ex]{\(\ast\ ) } } } : { { \text{\textbf{\textsl{set}}}}}/b & \longrightarrow & { { \text{\textbf{\textsl{set}}}}}/a \\ { } [ x\to b ] & \longmapsto & [ a\times_b x \to a ] . \end{aligned}\ ] ] in fibre notation , @xmath56 finally , also @xmath57[0ex]{\(\ast\)}}}$ ] in turn has a right adjoint , denoted _ lowerstar _ , which is more involved to describe synthetically , but whose description in terms of fibres is `` multiply along the fibres '' : @xmath58[0ex]{\(\ast\)}}}:{{\text{\textbf{\textsl{set}}}}}/a & \longrightarrow & { { \text{\textbf{\textsl{set}}}}}/b \\ ( x_a\mid a\in a ) & \longmapsto & ( \prod_{a\in a_b } x_a \mid b\in b )
. \end{aligned}\ ] ] if the categories @xmath59 and @xmath39 are replaced by the equivalent categories @xmath60 and @xmath48 , the three functors @xmath54 , @xmath57[0ex]{\(\ast\)}}}$ ] , and @xmath61[0ex]{\(\ast\)}}}$ ] can still be interpreted : @xmath57[0ex]{\(\ast\)}}}$ ] is just precomposition with @xmath62 .
its adjoints are left and right kan extension respectively ( cf . @xcite , ch .
x , for this notion ) . in this paper
we shall not need the explicit form of these functors .
although we actually formulate many results in terms of the functor categories @xmath48 , when it comes to functors we prefer to work with slices .
polynomial functors .
a diagram of sets @xmath63 & & e & \rto^p & & b & & \\ & \ldto^{s}&&&&&\rdto^{t } & \\ i & & & & & & & j \end{diagram}\ ] ] defines a _ polynomial functor _ @xmath64[0ex]{\(\ast\)}}}}\rto { { \text{\textbf{\textsl{set}}}}}/e \stackrel{p{_{\raisebox{-0.33ex}[-0.5ex][0ex]{\(\ast\)}}}}\rto { { \text{\textbf{\textsl{set}}}}}/b \stackrel{t{_!}}\rto { { \text{\textbf{\textsl{set}}}}}/j .\ ] ] in terms of indexed families and fibres , the formula for this polynomial functor is @xmath65 it is hence an @xmath66-indexed family of sums of products of @xmath67-indexed families .
particularly suggestive is the case @xmath68 , so that we are talking about a ` single polynomial in one variable ' . in this case
the formula boils down to @xmath69 morphisms.[nat ] morphisms of polynomial functors are just natural transformations
. one can show ( @xcite ) that a natural transformation @xmath70 between polynomial functors is uniquely represented by diagrams of the form @xmath71 by @xmath72 we denote the category of polynomial functors @xmath73 , and their natural transformations . for the manipulation of polynomial functors in terms of the representing sets , the following two facts are basic .
beck chevalley . given a pullback square a & ^ & b + + a & _ & b there are natural isomorphisms of functors @xmath74[0ex]{\(\ast\)}}}{\stackrel{\raisebox{0.1ex}[0ex][0ex]{\(\sim\ ) } } { \raisebox{-0.15ex}[0.28ex]{\(\rightarrow\)}}}{\varphi}{^{\raisebox{-0.25ex}[0ex][0ex]{\(\ast\)}}}\circ \beta{_!}\qquad \text { and } \qquad \beta{^{\raisebox{-0.25ex}[0ex][0ex]{\(\ast\)}}}\circ { \varphi}{_{\raisebox{-0.33ex}[-0.5ex][0ex]{\(\ast\)}}}{\stackrel{\raisebox{0.1ex}[0ex][0ex]{\(\sim\ ) } } { \raisebox{-0.15ex}[0.28ex]{\(\rightarrow\)}}}{\overline}{\varphi}{}{_{\raisebox{-0.33ex}[-0.5ex][0ex]{\(\ast\)}}}\circ \alpha{^{\raisebox{-0.25ex}[0ex][0ex]{\(\ast\)}}},\ ] ] usually called the _ beck chevalley
isomorphisms_. distributivity .
starting from maps @xmath75 , we can construct the following diagram by applying @xmath76[0ex]{\(\ast\)}}}$ ] to the map @xmath77 : ^[0ex][0ex]@xmath78_[-0.5ex][0ex]@xmath78a & ^ & _ [ -0.5ex][0ex]@xmath78a + + + & & & ^ & & & _ & & _ & & & + & & & & _ & & & & & & _ & + & ^&&&&&_&&_&&&&^ + & & & q&&&&&&&p & & & the composite @xmath79 is constructed as the top outline : start by taking the pullback at the common middle set , then lowerstar the result to arrive at what will be the top right - hand corner of the composite , and pull back to complete the distributivity pentagon . the diagram is completed by taking two more pullbacks as indicated .
the bridge diagram constructed is naturally isomorphic to the composite of the original functors by the beck
chevalley isomorphisms and distributivity .
the construction reflects closely how one composes two polynomial functions in elementary algebra : the key point is of course distributing the products involved in the outer polynomial over the sums of the inner .
the @xmath80-category of polynomial functors .
polynomial functors form a @xmath80-category - category ( see for example @xcite , ch .
xii ) is not essential to understand this work , but it is an efficient framework to set up some of the involved notions correctly . ] @xmath81 in which the objects are the slices of @xmath82 ( or equivalently , the categories @xmath83 for @xmath67 a set ) , the @xmath26-cells are the polynomial functors ( i.e. those isomorphic to one given by a diagram ) , and @xmath80-cells are arbitrary natural transformations .
@xmath81 is a strict @xmath80-category . by theorem 2.17 of @xcite
it is biequivalent to a bicategory whose objects are sets @xmath67 , whose @xmath26-cells are diagrams like and whose @xmath80-cells are diagrams like ( and where composition is described as in [ comp ] ) .
the theorem allows us to blur the distinction between bridge diagrams and the polynomial functors they represent , allowing for the conceptual benefit of the strict @xmath80-category @xmath81 and the computational benefit of the representing diagrams .
this is a characteristic aspect of the theory of polynomial functors .
we are here concerned with the aspect of polynomial functors as a categorification of the elementary algebra of polynomial functions . in this section we introduce a set - up to deal with ` rings ' of polynomial functors and their associated ` affine spaces ' , in the sense of algebraic geometry @xcite .
the results in this section appear for the first time , but they are not difficult .
one key issue is to impose the correct finiteness conditions .
certainly , our ` ground category ' should be the category of finite sets : all coefficients and exponents will now be required to be finite .
however , the polynomial rings we are interested in have infinitely many variables , so we can not just take the theory of polynomial functors internally to the category of finite sets .
. the category of finite sets , which we denote @xmath84 will be our coefficient ` ring ' .
more precisely , the monoidal operations of finite sums and products make it a distributive category @xcite . clearly
the set of isomorphism classes of @xmath7 is the set of natural numbers , and sum and products then yield the usual addition and multiplication of numbers . in short ,
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ @xmath6 is the burnside semiring of @xmath7 _ , _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ which is categorification in its purest form .
the notation @xmath7 stresses the algebraic aspect of @xmath85 , and we use this notation when we think of that category as the ground ring .
the same category also plays a geometric role , and as such we denote it @xmath86 leading to a natural notation for what plays the role of affine space .
namely , if @xmath67 is a set , then the functor category @xmath87 is the domain for the finite polynomial functors in @xmath67-many variables .
( note that @xmath88 is not equivalent to @xmath89 when @xmath67 is an infinite set : the latter is the category of maps @xmath90 with @xmath91 finite .
the former is equivalent to the category of maps @xmath90 with finite fibres . )
finite polynomials .
a polynomial @xmath92 is called _ finite _ when @xmath38 and @xmath91 are finite sets .
these are the polynomial functors on @xmath93 with values in finite sets .
we denote the category of these polynomial functors @xmath94 : = { \text{\textbf{\textsl{finpoly}}}}(i,1 ) .\ ] ] clearly @xmath95 $ ] is a distributive category and _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ the burnside semiring of @xmath95 $ ] is @xmath96 $ ] , the polynomial semiring in @xmath67-many variables , _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ as the notation also suggests .
moreover , _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ @xmath95 $ ] is the free distributive category on the set @xmath67 , _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ just as @xmath96 $ ] is the free commutative semiring on @xmath67 .
indeed , the category @xmath95 $ ] is generated freely under finite sums by the monomials , i.e. those @xmath97 for which @xmath98 .
from the characterisation of natural transformations in [ nat ] , we see that the category of monomial functors in @xmath67-many variables has arrows given by commutative triangles & & e + i & ( 2,1 ) & + & ( 2,1 ) & e , and so is equivalent to @xmath99 .
but it is well known that @xmath99 is the finite - product completion of @xmath67 . ` polynomial rings ' and ` affine space ' .
the distributive category @xmath95 $ ] is the category of finite polynomial functors on @xmath93 .
conversely , the category @xmath93 can be reconstructed from @xmath95 $ ] : define a _ character _ on a distributive category @xmath100 to be a functor @xmath101 preserving finite sums and finite products .
the form a category in which the morphisms are the finite - sum - and products - compatible natural transformations .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ @xmath93 is equivalent to the category of characters of @xmath95 $ ] . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ indeed , the category of sum - and - product preserving functors @xmath95\to{\mathbb{f}}$ ] is equivalent to the product - preserving functors @xmath102 , and since the domain is the product - completion of @xmath67 , we are finally left with the category of functors @xmath103 , which is what we denote @xmath93 .
appropriate maps : locally finite polynomial functors .
[ duality ] for a given set @xmath67 we now have the category @xmath95 $ ] playing the role of a polynomial ring , and the category @xmath93 playing the role of affine space .
we proceed to assemble these objects into @xmath80-categories @xmath104 and @xmath105 , respectively , which will be the ambient setting for defining the algebraic structures promised in the introduction .
the @xmath80-category @xmath104 is defined as follows .
its objects are the categories of the form @xmath95 $ ] ( i.e. free distributive categories on a set ) .
the @xmath26-cells are the finite - sum - and - product - preserving functors , and the @xmath80-cells are the compatible natural transformations ( i.e. those whose component on a finite sum is the sum of the components , and similarly with finite products ) .
the universal property of @xmath106 $ ] implies that the hom cats in @xmath104 are @xmath107,{\mathbb{f}}[i ] ) \simeq { \mathbb{f}}[i]^j .\ ] ] on the other hand , the @xmath80-category @xmath105 is defined as having the categories @xmath93 as objects , and as hom cats the categories of locally finite polynomial functors and their natural transformations , denoted @xmath108 : a polynomial functor @xmath109 is called _ locally finite _ when the maps @xmath46 and @xmath110 have finite fibres .
this means that only finite sums and finite products are involved .
it is not difficult to see that we have an equivalence of categories @xmath111^j .\ ] ] one may observe that the locally finite polynomial functors can also be charaterised as those @xmath112 such that for any @xmath0 belonging to @xmath106 $ ] , the composite @xmath113 belongs to @xmath95 $ ] , in analogy with the definition of regular maps in algebraic geometry .
almost tautologically we have an equivalence of @xmath80-categories @xmath114 justifying the symbols and the terminology . the tensor product and comonoidal structure .
the @xmath80-category @xmath115 has finite products : it is given by the equivalence of categories @xmath116 accordingly , the category @xmath104 has categorical sums , which we denote by @xmath117 .
all we need to know about it is the equivalence @xmath118 { \otimes}_{{\mathbb{f } } } { \mathbb{f}}[i_2 ] \simeq { \mathbb{f}}[i_1+i_2 ] , \ ] ] which is just dual to the previous display .
the neutral element for this tensor product is the category @xmath119 $ ] itself , which is an initial object in @xmath104 : the unique sum - and - product - preserving functor @xmath120 $ ] sends a finite set @xmath121 to the polynomial functor @xmath122 the objects in @xmath123 are algebras .
we would like to define bialgebras as comonoids in @xmath123 , or rather pseudo - comonoids ( i.e. coassociative only up to given coherent isomorphisms ) .
a bialgebra is then a category @xmath95 $ ] equipped with a comonoidal structure , i.e. functors @xmath124 \to { \mathbb{f}}[i]{\otimes}{\mathbb{f}}[i]$ ] .
every @xmath95 $ ] has two canonical such comonoidal structures , as we shall see in the next paragraph .
the main result of this paper establishes another comonoidal structure for the special case @xmath125 .
however , it is much more convenient to describe these structures on the other side of the duality , namely in @xmath126 , where the background structure is just the product , and we are talking about the familiar notion of monoidal structure instead of comonoidal structure .
canonical monoidal structures on @xmath93 .
fix a set @xmath67 .
the category @xmath93 has two canonical monoidal structures : pointwise sum and pointwise product .
these are given by locally finite polynomial functors .
the _ sum _ of @xmath127 and @xmath128 is the function @xmath129 which sends @xmath130 to @xmath131 . under the equivalence @xmath132
the sum map is polynomial : it is represented by @xmath133 the corresponding comonoidal structure on @xmath95 $ ] is given by @xmath134 & \longrightarrow & { \mathbb{f}}[i]{\otimes}_{\mathbb{f}}{\mathbb{f}}[i ] \\
x_i & \longmapsto & x_i{\otimes}1 + 1 { \otimes}x_i .
\end{aligned}\ ] ] here @xmath135 $ ] denotes the polynomial functor @xmath136 where @xmath137 picks out the element @xmath130 .
the neutral object for the sum is of course the constant empty set @xmath138 . under the equivalence @xmath139
, the corresponding map is represented by @xmath140 the corresponding counit for the comultiplication is given by @xmath141 .
the _ product _ of @xmath127 and @xmath128 is the function @xmath142 which sends @xmath130 to @xmath143 .
it is represented by @xmath144 the unit is the constant function @xmath26 .
the corresponding comonoidal structure on @xmath95 $ ] is given by @xmath145 and @xmath146 .
this categorifies the construction of the monoid - algebra on the free abelian monoid on @xmath67 . before coming to the promised new monoidal structure on @xmath9 in section [ sec : monoidal ] ,
we need some background on trees .
we briefly summarise the polynomial formalism of trees from @xcite .
trees.[polytree - def ] it was observed in @xcite that operadic trees can be conveniently encoded by diagrams of the same shape as polynomial functors .
we take this as the definition of tree : an _ operadic tree _ is a diagram of finite sets a & ^s & m & ^p & n & ^t & a satisfying the following three conditions : \(1 ) @xmath110 is injective \(2 ) @xmath147 is injective with singleton complement ( called the _ root _ and denoted @xmath26 ) . with @xmath148 ,
define the walk - to - the - root function @xmath149 by @xmath150 and @xmath151 for @xmath152 .
\(3 ) @xmath153 .
the elements of @xmath154 are called _
edges_. the elements of @xmath155 are called _
nodes_. for @xmath156 , the edge @xmath157 is called the _ output edge _ of the node . that @xmath110 is injective is just to say that each edge is the output edge of at most one node . for @xmath156 ,
the elements of the fibre @xmath158 are called _ input edges _ of @xmath42 .
hence the whole set @xmath159 can be thought of as the set of nodes - with - a - marked - input - edge , i.e. pairs @xmath160 where @xmath42 is a node and @xmath20 is an input edge of @xmath42 .
the map @xmath147 returns the marked edge .
condition ( 2 ) says that every edge is the input edge of a unique node , except the root edge .
condition ( 3 ) says that if you walk towards the root , in a finite number of steps you arrive there .
the edges not in the image of @xmath110 are called _
leaves_. from now on we just say _ tree _ for ` operadic tree ' . the tree @xmath161 is the _
trivial tree _ , which we denote by .
cartesian morphisms ( cf .
@xcite ) . [ cart ]
a _ cartesian morphism _ of polynomial endofunctors is by definition a diagram @xmath162 i ' & \lto & e ' \sepbk & \rto & b ' & \rto & i ' \\
\dto<\alpha & & \dto & & \dto & & \dto>\alpha \\
i & \lto & e & \rto & b & \rto & i .
\end{diagram}\ ] ] if @xmath163 is an identity map , then this corresponds precisely to cartesian natural transformations between polynomial endofunctors ( i.e. whose naturality squares are cartesian ) . in the general case ,
the cartesian morphisms like are outside the scope of the @xmath80-category @xmath81 , but they can be reduced to it by some base change @xcite .
morphisms of trees ( cf .
1.1).[sub ] a _ morphism of trees _ is by definition a cartesian morphism of the associated polynomial endofunctors .
hence edges go to edges , and the cartesian condition means that a node of arity @xmath164 is mapped to a node of arity @xmath164 .
the morphisms of trees are also called _ tree embeddings _ since in fact each of the components of such a map is automatically injective : indeed , the tree axioms above imply that a morphism necessarily commutes with the walk - to - the - root function @xmath165 , and together with arity perservation this forces the maps to be injective .
hence the category of trees and tree embeddings , denoted @xmath166 , is mostly concerned with subtrees , but note that it also contains automorphisms of trees .
a tree embedding is _ root - preserving _ when it sends the root to the root . in formal terms , these are diagrams such that also the left - hand square is cartesian , @xcite , 1.1.13 . an _ ideal embedding _ ( or an _ ideal subtree _ ) is a subtree for which every descendent edge and node is also in the tree , @xcite , 1.1.9 .
there is one ideal subtree generated by each edge in the tree .
the ideal embeddings are characterised as having also the right - hand square of cartesian .
ideal embeddings and root - preserving embeddings admit pushouts along each other in the category @xmath166 , @xcite , 1.1.20 .
the most interesting case is pushout over a trivial tree : this is then the root of one tree and a leaf of another tree , and the pushout is the grafting onto that leaf .
decorated trees : @xmath0-trees.[ptree ] let @xmath0 denote a finitary polynomial endofunctor on @xmath167 , represented by @xmath168 by definition ( @xcite , 1.2 ) , a _
@xmath0-tree _ is a diagram a&&m & & n & & a + & & & & & & + i&&e&&b & & i . where the top row is a tree .
unfolding the definition , we see that a @xmath0-tree is a tree whose edges are decorated in @xmath67 , whose nodes are decorated in @xmath38 , and with the additional structure of a bijection for each node @xmath169 ( with decoration @xmath170 ) between the set of input edges of @xmath164 and the fibre @xmath171 , subject to the compatibility condition that such an edge @xmath172 corresponding to @xmath173 has decoration @xmath174 , and the output edge of @xmath164 has decoration @xmath157 .
note that the @xmath67-decoration of the edges is completely specified by the node decoration together with the compatibility requirement , except for the case of a trivial tree .
the free monad on a polynomial endofunctor .
( for the notion of monad , see @xcite , ch.vi , or @xcite , ch . 10 . )
by a _ polynomial monad _ we mean a monad in the @xmath80-category of polynomial functors and cartesian natural transformations , cf .
the forgetful functor from polynomial monads to polynomial endofunctors has a left adjoint , the free - monad functor .
it has the following pleasing description in terms of trees , cf .
@xcite , prop .
1.2.8 : the free monad on a polynomial endofunctor @xmath0 is the polynomial monad represented by @xmath175 where @xmath176 is the set of ( isomorphism classes of ) @xmath0-trees , and @xmath177 is the set of ( isomorphism classes of ) @xmath0-trees with a marked leaf .
the left - most map returns the decoration of the marked leaf ; the right - most map returns the decoration of the root , and the middle map @xmath178 just forgets the mark .
the monad structure is given by grafting of trees , an operation that allows an easy and completely formal description in the category of trees in terms of pushouts , cf .
@xcite , prop .
examples of the notion of @xmath0-trees , for suitable choices of @xmath0 , are planar and binary trees , or more specialised examples like trees with nodes decorated by primitive 1pi graphs of a quantum field theory @xcite , or the opetopes of higher category theory @xcite .
the polynomial endofunctor @xmath0 will be held fixed throughout , and from now on we put @xmath179 .
we let @xmath180 denote the free monad on @xmath0 .
we shall need an explicit description of the monad structure of @xmath180 .
as in [ comp ] , the composite @xmath181 , is built by the following diagram , whose constituents we now make explicit : & & & & & & &&&&^p & & & = q_[-0.5ex][0ex]@xmath78(_i ) + & & & & & & & & _ + & & & & & & & _ i &&&&>r + & & & & & & & & & & & + & & &&_q&&&&&&_q & + & & & & & & & & & & & & + i&&&f&&&&i&&&f&&&i the construction starts from the bottom by forming the central pullback square : @xmath182 is the set of pairs @xmath183 such that the root of @xmath121 is of the same type as the marked leaf of @xmath21 , in other words a simple grafting of one tree onto another .
this can also be interpreted as the set of trees with one marked edge : the two projections are then : return the ideal subtree generated by that edge ( that s @xmath121 ) , and return the root - preserving tree obtained by pruning that ideal subtree ( that s @xmath21 ) .
next we compute @xmath184[0ex]{\(\ast\)}}}({\mathbf{t}}\times_i { \mathbf{t}}')$ ] .
( note that this set can also be described as @xmath185 . ) by definition of the lowerstar operation , the result is a set over @xmath10 whose fibre over a fixed tree @xmath186 is the set of maps from the set of leaves of @xmath21 to the set of all trees with matching root .
this data is equivalent to the grafting of all those leaf - indexed trees onto the leaves of @xmath21 .
alternatively this data amounts to the inclusion of @xmath21 into the big tree resulting from the graftings .
in other words , @xmath187 is the set of trees with a cut , and the map @xmath188 returns the tree found below the cut .
the pullback along @xmath178 is the set @xmath189 of pairs consisting of a marked tree and graftings onto all leaves , which amounts to trees with a pointed cut ; the map @xmath190 just forgets the mark in the cut .
the evaluation map @xmath191 is described like this : if we think of the elements of @xmath189 as given by a tree together with a marked leaf and a map from each leaf to @xmath10 ( with matching root ) , then the map @xmath192 simply applies that map to the marked leaf hence obtaining a single tree with correct root , and hence an element in @xmath182 .
( we do nt really need the remaining left - hand part of the diagram , but here it is , for completeness : the upper left - hand corner is the set of trees with a pointed cut and a marked leaf in the subtree corresponding to the point .
( since that marked leaf together with the cut automatically provides a point on the cut it is enough to say : tree with a cut and a marked leaf . ) the set just below it is the set of trees with a marked edge and a marked descendant leaf . )
finally , we describe the structure maps of the free monad @xmath180 : the multiplication @xmath193 is the cartesian natural transformation ff : & & i & & & & & & i + & & & & & & > m & & + f : & & i&& & & & & i . where the maps in the middle simply forget the cut .
the unit for the monad , @xmath194 is given essentially by the map @xmath195 assigning to each ` colour ' @xmath196 the trivial tree with edge of colour @xmath130 .
we summarise the maps @xmath197 , @xmath46 , @xmath62 , @xmath198 , for later use : & & & & & ^p&&=f ( ) + & ( 0,0)(-20,16 ) ( 45,45)(-30,30)(-45,-45 ) ( -24,20)(0,0)[t]@xmath199 ( -45,-45)(-1,-4)0 & & + & & _ i &&&&>r + & & & & & & & ( 0,0)(6,16 ) ( -12,87)(30,0)(-15,-95 ) ( 16,10)(0,0)[t]@xmath200 ( -14.5,-95)(-1,-4)0 + & & & & & & + & & & & & & & + & & i&&&f&&&i + & & & & & & & + & & & & & &
in this section the symbol @xmath10 stands for the set of isomorphism classes of @xmath0-trees , for any fixed finitary polynomial endofunctor @xmath0 ( and similarly for the decorated symbols , @xmath187 , etc . ) .
the construction of the monoidal structure works also for abstract trees , but then the various sets and maps no longer come from the free - monad construction , and need to be defined in a more ad hoc manner .
we discuss this in the next section . theorem . _ the polynomial functor @xmath201 defined by _ & & + & ^p , & & & & + & ^f+r&&&&&^m & + + & & & m & & & & is a monoidal structure on @xmath9 . its unit is the functor @xmath202 given by & & 0 & & i & & + & & & & & ^e & + 0 & & & u & & &
. description of the maps
. an interesting feature of this monoidal structure is that all the involved maps are recognised as those occurring in the polynomial description of the free - monad construction .
we review the definitions of the sets and maps : * @xmath10 is the set of iso - classes of @xmath0-trees . if @xmath180 denotes the free monad on @xmath0 , then @xmath203 .
* the set @xmath187 is the set of iso - classes of @xmath0-trees with a cut .
it appears as @xmath204 .
more formally , the set of iso - classes of trees with a subtree containing the root . *
the map @xmath205 is the multiplication of the monad , @xmath206 , i.e. forget the cut . *
the set @xmath189 is the set of iso - classes of @xmath0-trees with a pointed cut , and @xmath46 is the map that forgets the point .
more formally , @xmath207[0ex]{\(\ast\)}}}{\overset { { \begin{picture}(8,1)(0,0 ) \put(0,0){\line(1,0){8 } } \end{picture } } } { { \mathbf{t } } } } , \ ] ] where @xmath208 is the set of iso - classes of @xmath0-trees with a marked leaf ( it appears as the total space of @xmath209 in the diagram representing @xmath180 ) , and @xmath210 is the projection . *
the map @xmath211 returns the @xmath0-tree below the cut . *
@xmath212 returns the ideal subtree generated by the edge marked by the pointed cut .
this is the most complicated map to explain formally , but it occurs already in the diagram for @xmath181 , and is composed of the @xmath10-component of the counit of the @xmath213[0ex]{\(\ast\)}}}{\dashv}q{_{\raisebox{-0.33ex}[-0.5ex][0ex]{\(\ast\)}}}$ ] adjunction ( that s the evaluation map @xmath214[0ex]{\(\ast\)}}}{\overset { { \begin{picture}(8,1)(0,0 ) \put(0,0){\line(1,0){8 } } \end{picture } } } { { \mathbf{t}}}}\to { \mathbf{t}}\times_i { \mathbf{t}}'$ ] ) followed by the projection to @xmath10 . *
finally , @xmath195 assigns to each ` colour ' @xmath196 the trivial tree with edge of colour @xmath130 .
note that here and throughout , the name of a set is also used to indicate its identity map .
so for example @xmath215 denotes the map whose first component is @xmath46 and whose second component is the identity map on @xmath187 .
note that the polynomial is a ` sum - wise tensor product ' of a non - linear functor ( in the first set of variables ) : & & & ^p & & & & + & ^f&&&&&^m & + & & & & & & & ( which concerns taking all cuts and retaining the forest of cut - off branches ) , and a linear functor ( in the second set of variables ) : & & & ^= & & & & + & ^r&&&&&^m & + & & & & & & & ( which is about taking all cuts and retaining the bottom tree , which of course is a single tree , hence the linearity ) .
proof of associativity . in the following diagram , which commutes strictly , the left - hand vertical polynomial is @xmath216 , and the bottom is @xmath12 .
similarly the top polynomial is @xmath217 and the right - hand vertical polynomial is @xmath12 . we show that both composites are naturally isomorphic to the polynomial indicated by the diagonal ( to be detailed in the proof ) . in both cases
this amounts to performing the constructions as in [ comp ] .
this natural isomorphism provides the associator , which is part of the monoidal structure .
( the pentagon equation for the associator is not established explicitly .
we invoke instead a general coherence principle : since the associator is constructed by canonical isomorphisms ( beck chevalley and distributivity ) , it is coherent . )
+ + & & ^&&(+)+&^&+&^&+ + & ( 4,4)^ & & & & & > + & & + + ( + ) & ^&&^&&^ & & & > + & & + + & _ & & _ & & _ & & ^ & + + + & _ & + & & _ & & & _ & .
we start with the lower left - hand composite .
the first step is to take a pullback in the lower left - hand corner : + & ^&+ + + + & _ & + .
here @xmath218 is the set of trees with two non - crossing cuts , or more formally , a sequence of two root - preserving inclusions @xmath219 .
the upper cut is the one coming from the lower right - hand corner of the diagram , and the lower cut comes from the upper left - hand corner of the diagram .
hence the map denoted @xmath220 forgets the lower cut .
the right - hand map @xmath221 we shall now lowerstar along @xmath222 to complete the distributivity pentagon : + & & & ^ & + + + & & _ & & .
the resulting right - hand map is just @xmath220 . indeed ,
the lowerstar operation amounts to multiplying along the fibres .
but since in the left - hand summand the map we lowerstar is just the identity of @xmath223 , no contribution comes from this summand , and in the right - hand summand we are lowerstarring along the identity map , hence the result is just @xmath220 , the map that forgets the lower cut . we have now shown that the set @xmath218 appears at the crucial point of the diagonal polynomial when constructed from the lower left - hand side .
we proceed to construct it also from the upper right - hand side of the big diagram .
the argument is different because of the non - symmetry of the tensor product .
the pullback square in the upper right - hand corner is , typographically transposed : + & ^ & + + + + & _ & + .
here there are two new symbols : @xmath224 denotes the set of trees with a marked cut and a further cut in the tree above the mark .
the map @xmath225 forgets the ` short ' cut .
we shall now lowerstar the map @xmath226 along @xmath227 , to complete the distributivity pentagon + & & & ^ & + + + & & _ & & .
the claim is that the right - hand map is just @xmath228 , the map that forgets the upper cut .
we compute this fibre - wise over an element @xmath229 in @xmath230 .
the @xmath231-fibre has one element for each leaf @xmath20 of @xmath21 . for each leaf @xmath20 ,
the @xmath225-fibre consists of the possible cuts in the ideal subtree @xmath232 generated by @xmath20 .
lowerstarring means multiplying these fibres , so it amounts to giving a cut in @xmath233 for each leaf @xmath20 of @xmath21 .
altogether , this amounts to giving a total cut in @xmath13 above the original cut @xmath33 .
this is once again the set @xmath218 , but the projection @xmath234 this time forgets the upper cut .
( note that there is no contribution from the right - hand summand , as it has trivial fibres . ) finally , note that the square appearing where the two constructions meet , & ^ & + + & _ & commutes , since it amounts to forgetting first the upper cut and then the lower , or the other way around .
the remaining part of the big diagram is only about taking pullbacks , and presents no difficulties .
proof of the unit axiom .
the unit is the functor @xmath202 represented by @xmath235 where the last map associates to an edge colour the corresponding trivial tree . that this is the unit object means that when left - added to the identity functor on @xmath10 , @xmath236 and composing with @xmath12 yields the identity functor . and similarly of course with right - adding the identity functor .
checking this amounts to filling the following big diagram : & & ^&&&^&+i&^&+ + & ( 4,4 ) & & & & & > + & & + & ^&&^&&^ & & & > + & & + i+&_&&_&&_&&^ & + + + & _ & + & & _ & & & _ & . as with associativity , we content ourselves to checking that the two constructions meet at the distributivity squares .
starting at the lower left - hand corner with a pullback : i+&^f+r&v+ + + + & _
f+r&+ the set @xmath237 appearing is the set of trees with a marked cut such that the mark itself is a leaf ( in other words , the upper tree is a trivial tree ) . at random . ]
the map @xmath238 is the inclusion of these special pointed cuts into the set of all pointed cuts .
now lowerstar along @xmath239 , to complete the distributivity pentagon : v+ & & & ^ & w&= + + + & & _ p , & & .
this amounts to requiring for each mark of the cut , that the tree above it is trivial ; in other words it is the set @xmath240 of cuts that only take leaves , i.e. the set of trivial root - preserving inclusions , i.e. the set @xmath10 itself .
( the map @xmath241 is of course the inclusion of this particular kind of cuts into the set of all cuts . )
the check is similar for the right - hand unit axiom , except that the set in the corner is the set of trees with the root cut , which again is naturally identified with @xmath10 itself .
bialgebra of polynomial functors . by the duality of [ duality ]
, the monoidal structure on @xmath9 induces a comonoidal structure on @xmath8 $ ] , & ^ & [ ] & ^ & [ ] _ [ ] : the comultiplication is simply defined by precomposition with @xmath12 and the counit is defined by precomposition with @xmath242 . in detail , if @xmath243 $ ] is a polynomial functor in @xmath10-many variables , @xmath244 we compute @xmath245 as the composite & & + & ^p , & & & & & & e & & b & & + & ^f+r&&&&&^m & & & & & & & + + & & & & & & & & & & & & & 1 .
note that the comonoidal structure is automatically multiplicative , for formal reasons : multiplication in @xmath8 $ ] is dual to the diagonal functor @xmath246 , and every monoidal structure on @xmath9 is compatible with this diagonal .
it is instructive to calculate @xmath247 on a multiplicative generator of @xmath8 $ ] : these are the single - variable polynomials @xmath248 where the map @xmath249 picks out the tree @xmath250 .
the composition in this case just amounts to taking fibres , so the result is & & ( ) _ t+()_t & ^ p_t,()_t & & ( ) _ t & & + & ^f+r & & & & & & + + & & & & & & & 1 .
here @xmath251 is the set of cuts in the specific tree @xmath13 , and the fibre over that is the a two - component set where the left - hand component is the set of all pointed cuts on @xmath13 , and the right - hand component is again the set of cuts on @xmath13 . spelling out what the polynomial does as a functor
, we see that it is exactly the functorial version of the formula from [ comultmy ] : we have to sum over the set @xmath251 ( that s the set of cuts ) , then multiply the fibres for each cut @xmath252 ; the fibre has an element for each edge in the cut ( and one extra ) : for each edge we must apply @xmath62 , which gives the branch factor of @xmath16 , and multiply all these , getting altogether what is denoted @xmath16 , and for the extra point of the fibre we must apply @xmath197 , which gives the tree @xmath17 below the cut , which is to be placed in the right - hand factor .
the counit associates to each tree @xmath13 a polynomial in zero variables , i.e. just a finite set .
if the tree is represented as @xmath253 , we need to compose & & 0 & & i & & & & 1 & & 1 & & + & & & & & & & & & & & & + 0 & & & & & & & & & & & & 1 .
the result is & & 0 & & & i _ 1 & & + & & & & & & & + 0 & & & & & & & 1 , i.e. the set @xmath254 , which is the singleton set @xmath26 if @xmath13 belongs to @xmath67 and the empty set @xmath138 otherwise .
again it is clear that this is the functorial analogue of the specification in section [ sec : bialg ] .
trees ( irrespective of any @xmath0 ) . we have worked with @xmath0-trees for two reasons :
the first is to allow the connection with the free - monad construction .
the second is the fact that @xmath0-trees ( in the set - based setting we are working in ) are rigid .
this guarantees that taking sets of isomorphism classes is well - behaved .
abstract trees , in the sense of [ polytree - def ] , are not @xmath0-trees for any polynomial endofunctor over @xmath82 .
the ` terminal endofunctor ' in the category of polynomial endofunctors and their cartesian morphisms which should make the abstract trees @xmath0-trees does not exist . also ,
trees clearly may have non - trivial automorphisms , so taking sets of isomorphism classes may be prone to errors , and care is needed .
reinterpretation of the sets and maps in section [ sec : monoidal ] . in spite of the above remarks , it is nevertheless possible to mimic all the constructions of the previous section ,
if just the involved sets and set maps are defined a little bit differently .
the starting point is the same : we now let @xmath10 denote the set of isomorphism classes of abstract trees .
hence the basis for the bialgebra is the same as used already in section [ sec : bialg ] .
but while for @xmath0-trees , the set @xmath187 can be described as isomorphism classes of root - preserving inclusions , this description does not work for abstract trees .
indeed , it is crucial that the fibres of the projection @xmath255 have the correct cardinality : for example , the abstract tree certainly has five different cuts , but two of them are isomorphic as abstract cuts .
so instead of defining @xmath187 as the set of isomorphism classes of cuts , it is necessary first to choose a representative for each iso - class in @xmath10 , and then for each such representative @xmath13 define @xmath256 to be the set of cuts of that specific tree @xmath13 , and finally let @xmath187 be the disjoint union of all those fibres .
similar care is needed to define @xmath208 and @xmath189 , and the other symbols involved .
the fact that all the symbols are defined fibrewise over @xmath10 , instead of being defined abstractly in terms of isomorphism classes , guarantees that all the pullback and lowerstar operations ( which are fibrewise operations ) yield the expected results , and one can check that the proof of associativity and the unit axiom work equally well for abstract trees , with these provisos .
groupoids instead of sets .
clearly the remedies just explained are rather ad hoc . a better and more conceptual way to account for abstract trees and @xmath0-trees on equal footing consists in upgrading the theory from sets to groupoids .
this can be seen as one further step of categorification .
first of all it is necessary ( and possible ) to upgrade the theory of polynomial functors from sets to groupoids : this means that the representing diagrams for polynomial functors @xmath257 should then be ( suitably homotopically finite ) groupoids , and the functors themselves go between slices of the category of groupoids ( slices being taken in the homotopical sense ) . with
the appropriate adjustments , everything works as for sets : pullbacks and fibres have to be homotopy pullbacks and homotopy fibres , if the maps are not groupoid fibrations ; sums are replaced by slightly fancier colimits , and everything is up to equivalence of groupoids instead of isomorphism of sets .
this theory correctly incorporates all questions of symmetries of objects @xcite . in this setting , abstract trees are @xmath91-trees for the ` exponential functor ' @xmath258,\ ] ] represented by the groupoid diagram @xmath259 where @xmath260 is the groupoid of finite sets and bijections , and @xmath261 the groupoid of finite pointed sets and basepoint - preserving bijections , and where the quotient is the weak groupoid quotient ( sewing in paths instead of collapsing ) .
dolan @xcite for further discussion of such constructions , and the exponential functor in particular . in the groupoid setting , one works with the groupoid of trees , instead of with its set of isomorphism classes .
the practical benefit , in the presence of automorphisms of objects , is the same as working with moduli stacks rather than coarse moduli spaces in algebraic geometry .
the choice in this paper to work with sets is first of all that this theory is already available .
second , it is preferable to expose the essential ideas and constructions in the transparent rigid set - up , so as not to burden the arguments with homotopy theory .
however , to ease the eventual upgrade to groupoids , the exposition is careful to perform the arguments in a clean , functorial language , favouring natural isomorphisms over element - based constructions .
in fact , the theory of polynomial functors is currently being worked out for infinity - groupoids , in joint work with david gepner .
one important insight is that @xmath262-groupoids play just the same role for @xmath262-categories as sets play for categories .
seeing how the theory goes for sets is therefore an important first step for the more general case of groupoids and @xmath262-groupoids . .
. in _ homotopy methods in algebraic topology ( boulder , co , 1999 )
271 of contemp .
math . , pp .
soc . , providence , ri , 2001 .
http://arxiv.org/pdf/math-ph/9907010[arxiv:math-ph/9907010 ] . | we exhibit a monoidal structure on the category of finite sets indexed by @xmath0-trees for a finitary polynomial endofunctor @xmath0 .
this structure categorifies the monoid scheme ( over @xmath1 ) whose semiring of functions is ( a @xmath0-version of ) the connes kreimer bialgebra @xmath2 of rooted trees ( a hopf algebra after base change to @xmath3 and collapsing @xmath4 ) .
the monoidal structure is itself given by a polynomial functor , represented by three easily described set maps ; we show that these maps are the same as those occurring in the polynomial representation of the free monad on @xmath0 .
joachim kock * keywords : * rooted trees , hopf algebras , categorification , monoidal categories , polynomial functors , finite sets .
* mathematical subject classification : * 05c05 ; 16t05 ; 18a99 . |
the solar corona is intricately structured magnetically , and hosts myriads of magnetic arcades , loops and twisted flux ropes .
flux ropes are vital ingredients in many dynamical models of solar eruptions , where their internal as well as external magnetic field variation can render them unstable to kink @xcite , torus @xcite , or even tilt - kink @xcite ideal mhd instability routes . twisted flux ropes , and the concave upward parts of their field topology , are also the preferred sites to host solar filaments . solar filaments ( also known as prominences when observed above the limb ) are large masses of relatively cool ( @xmath0 k ) and dense ( about @xmath1 @xmath2 ) plasma suspended in the hot corona @xcite .
although the filaments have been studied since the end of the xix century , their formation and stability are not fully understood yet ( e.g. , * ? ? ?
* ; * ? ? ?
many filaments end up with a final eruption ; the filament eruptions are often associated with the coronal mass ejections @xcite , although it is not clear which of these phenomena is primary and which one is secondary . in addition
, filament eruptions in active regions can play an important role in the initiation and development of , at least , some types of solar flares @xcite .
solar flux ropes and the filaments / prominences that reside within them show distinctive emission at radio frequencies .
in particular , the microwave emission bands ( frequencies of about @xmath3 ghz ) are well covered by the current and upcoming radio observing facilities .
radio views at these frequencies allow us to study the free - free emission and absorption properties in the solar flux ropes and filaments , or even identify gyrolayers in sufficiently strong magnetic field concentrations .
flux ropes , with or without embedded prominences , can exist in quiescent as well as in active solar regions , and can thus be characterized by very different thermodynamic and magnetic properties . in this paper
, we use magnetohydrodynamic models of solar flux ropes to generate synthetic radio images at centimetric radio wavelengths .
simulating the thermal bremsstrahlung and gyroresonance emissions requires the knowledge of density , temperature and magnetic field variation throughout the flux rope structures . in practice , this requires using flux rope models that go beyond simplified force - free , or zero plasma beta assumptions , which by construction ignore the thermodynamic variation altogether .
nevertheless , zero - beta modeling is able to provide insight into the dominant flux rope dynamics , e.g. , when studying sympathetic eruptions due to removal of overlying stabilizing flux in multiple flux rope configurations @xcite , or in reproducing the slow rise and partial eruption phase of split filament systems using a double flux rope configuration @xcite .
the same zero beta paradigm has been used to study how magnetic field arcades can be deformed into helical flux ropes by means of flux cancellation on the solar surface @xcite . in this work
, we use a modern variant of these early studies , where either isothermal , finite - beta mhd , or full thermodynamically consistent mhd simulations provide density - temperature - magnetic field conditions that represent force - balanced configurations .
a brief survey of current observational findings on the solar filaments / prominences and their filament channels as seen at radio wavelengths is presented in section [ s - review ] .
the mhd models are discussed in section [ s - mhd ] , and the way to generate synthetic radio views from them is presented in section [ s - radio ] .
the obtained radio views on the virtual flux ropes and filaments are presented in sections [ res - isothermal ] and [ res - filament ] , respectively .
the conclusions and outlook for future work are given in section [ conclusion ] .
the filaments can be observed in a broad range of wavelengths . in the h@xmath4 spectral line , they are seen as bright structures ( prominences ) above the limb or as dark filaments when observed against the solar disk . in the extreme ultraviolet ( euv )
emission , the appearance of the filaments is wavelength - dependent as it reflects the formation temperatures of the euv lines @xcite . in radio observations ,
the filaments were firstly detected at millimetric and centimetric wavelengths @xcite , and later at lower and higher frequencies as well @xcite . to date , the most comprehensive observations have been performed in the microwave range using the nobeyama radioheliograph and siberian solar radio telescope , with a particular focus on the eruptive processes ( * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* etc . ) . the dominant mechanism of generation / absorption of the radio emission in filaments is thermal bremsstrahlung ( free - free emission and absorption ) ; in addition , dense cores of the filaments block propagation of radio emission at the frequencies below the local plasma frequency .
as a result , the appearance of the filaments in the microwave range ( at the frequencies of @xmath5 ghz ) is usually similar to that in h@xmath4 : they look like dark structures ( brightness depressions ) when observed on the solar disk and like bright prominences when observed above the limb . at the same time
, the quantitative characteristics of the microwave emission from the filaments depend on many factors and , first of all , on the considered emission frequency . at the frequencies of
about @xmath3 ghz the filaments are expected to be optically thick , i.e. , their observed brightness temperatures @xmath6 should be nearly equal to their kinetic temperatures @xmath7 .
this has been confirmed by observations : the typical observed brightness temperatures of the filaments in the mentioned frequency range are of order of @xmath8 k both for the filaments on the disk and above the limb ( * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* etc . ) .
when multi - frequency observations of a particular filament are available , the brigthness temperature seems to be independent on the frequency .
the filaments on the disk look dark because they absorb the emission from the underlying hotter layers of the solar atmosphere ( the quiet sun emission ) .
since the brightness temperature of the quiet sun @xmath9 decreases with frequency , the on - disk filaments at lower frequencies , as a rule , demonstrate a higher contrast @xcite .
another important factor is the filament temperature : e.g. , at the frequency of 17 ghz the filaments on the solar disk become invisible when heated to @xmath10 k ( the quiet sun temperature at this frequency ) , although they remain visible ( and even become brighter ) above the limb @xcite . at higher frequencies ,
the free - free emission / absorption mechanism becomes less effective and at the frequencies of @xmath11 ghz the filaments can become optically thin . as a result , we obtain @xmath12 for the filaments above the limb and @xmath13 ( partial absorption of the quiet sun emission ) for the filaments on the solar disk . in this frequency range ,
the observed brightness temperatures of the filaments above the limb vary from the typical filament temperatures of @xmath14 k to much lower values .
similarly , the observed filaments on the disk vary in appearance from well - defined dark stripes to barely noticeable brightness depressions ( relative to the quiet sun level ) .
both the brightness temperature of the filaments above the limb and the relative brightness depression of the filaments on the disk tend to decrease with the frequency ( see * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* etc . ) .
the frequency dependence of the observed brightness temperature can be used to diagnose the temperature and density of the filaments ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
since the angular resolution in the radio range is in most cases insufficient to resolve directly the filament widths , these widths are estimated using deconvolution or forward - fitting when the observations are compared with the model images convolved with the instrument beam .
the filament widths inferred from the microwave observations are variable : sometimes they are nearly the same as the widths observed in h@xmath4 ( * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* etc . ) , while sometimes they exceed the widths of the h@xmath4 filaments by a factor of about @xmath15 ( * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* etc . ) ; the observed widths of the radio filaments tend to decrease with the frequency .
most likely , this is caused by inhomogeneous structure of the filaments , since the radio emission can be affected not only by the dense and cool filament core , but also by the transition region between the filament and the surrounding corona , where the plasma density gradually decreases and the temperature increases ( e.g. , * ? ? ?
* ) ; this transition in different events can be more or less sharp .
the previous section summarizes the microwave observations of well developed filaments with a cool dense core ; the low temperature is the key factor affecting appearance of these filaments in the observations . at the same time
, the hot overdense flux ropes ( i.e. , with the same temperature as the surrounding coronal plasma ) are also likely to exist . for example , in the study of @xcite evolution of a helical magnetic flux rope simulated in isothermal mhd model have led to concentration of plasma and magnetic field within the flux rope .
such simulations confirm that overdense flux ropes can be formed even in completely isothermal models , i.e. , the isothermal approximation is well justified during most stages of the flux rope evolution ; a nearly constant temperature is maintained due to the fast thermal conduction in the lower solar corona and slow change of magnetic field there . on the other hand , formation of a cool filament requires development of the thermal instability ; once the threshold for this instability is reached ( which can take a long time ) , the filament cooling happens very quickly @xcite .
therefore we conclude that the isothermal ( or nearly isothermal ) flux ropes with enhanced plasma density and magnetic field strength may be a common phenomenon in the solar corona .
the forming filament channels without filaments inside can be difficult to identify in observations . since all filaments are located above the neutral lines of the photospheric magnetic field ( although not all neutral lines give rise to filaments ) , we can expect some specific features associated with the neutral lines .
indeed , the so - called neutral - line - associated microwave sources ( nls ) are a common phenomenon ( see , e.g. , the papers of * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* and references therein ) .
as the name implies , nls occur in the vicinity of magnetic neutral lines in active regions ; they are bright , compact and relatively stationary ( can live up to several days ) .
the nls have typical brightness temperatures of a few mk ( up to @xmath16 mk ) .
the spectral observations reveal a sharp spectrum with the peak at @xmath17 ghz and a steep decrease at higher frequencies @xcite ; nevertheless , the brightness temperature can reach @xmath18 mk at the frequencies of @xmath19 ghz as well @xcite .
such spectra can not be explained by the free - free emission mechanism ; they are consistent with the thermal gyroresonance emission in a very hot plasma ( and strong magnetic field ) or with the nonthermal gyrosynchrotron emission .
the physical nature and formation mechanism of the nls are yet unknown . according to the most popular explanation
, the emission is produced in the arcade of magnetic loops overarching the filament channel ( * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* etc . ) ; this model is supported by observations of the nls associated with the h@xmath4 filaments ( * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* etc . ) . on the other hand ,
the thermal free - free and/or gyroresonance emission from the flux rope ( which is characterized by high plasma density and strong magnetic field ) can make a significant or even dominant contribution .
@xcite have concluded that the magnetic field strength in the nls ( estimated using the photospheric magnetograms and their force - free extrapolations ) is insufficient to produce the gyroresonance emission at the observed ghz frequencies ; the nonthermal gyrosynchrotron emission mechanism has been proposed ( although it requires a stable long - living source of accelerated electrons ) .
however , we note that the force - free approximation is essentially inapplicable to the flux ropes .
as will be shown below , the mhd processes driven by the footpoint motions of a magnetic arcade can result in formation of strong magnetic fields in the corona .
obviously , making more definite conclusions about the nature of nls and their relation to the filament channels requires observations with higher spatial resolution than now .
in the remainder of this paper , we use three - dimensional magnetohydrodynamic ( mhd ) models of flux ropes , without and with actual filaments embedded , to create virtual radio views at varying radio frequencies .
we should note that the mhd simulations are computationally expensive and take a long time ; the choice of the simulation parameters is restricted by the technical considerations as well .
therefore we have studied only a limited number of mhd models , which illustrate different regimes of the flux rope formation .
the model parameters have not been chosen to fit any particular observed event , but represent several general cases ; therefore , a direct quantitative comparison of our results with observations would be difficult .
nevertheless , our results demonstrate qualitatively how the flux rope evolution affects the radio emission , and how the observable features are linked to the different source parameters .
the isothermal mhd models of helical magnetic flux ropes follow the approach presented in @xcite , where they demonstrated the formation process of a large - scale flux rope in a quiet sun region with finite plasma beta from an initially bipolar arcade .
the initial arcade was extrapolated as linear force - free field from an artificial photospheric magnetogram .
the deformation from a pure arcade to a flux rope configuration with overarching arcade was implemented by a combination of photospheric vortical and converging flow patterns enforced by the bottom boundary prescriptions .
this ultimately led to a flux rope configuration with the magnetic field and plasma concentration within the flux rope . our first model ( the `` low - field '' model corresponding to a quiet - sun filament channel ) is the same as in the study of @xcite .
in this model , the photospheric magnetogram had the extrema of @xmath20 g , the computational box dimensions were 240 mm ( length ) , 180 mm ( width ) , and 120 mm ( height ) , and the spatial grid resolution was 0.47 mm .
the simulations were performed until the maximum plasma velocity in the domain dropped below a certain threshold ( 3.8 km @xmath21 , see * ? ? ?
* ) , which was considered as the endstate .
the evolution of the magnetic field and plasma in this model is shown in fig . 4 in the paper of @xcite .
the resulting flux rope configuration is characterized by a low magnetic field strength typical for quiescent filament conditions : the vertical field component at the bottom of our domain , which is located at 3 mm above the photosphere , attains values of up to @xmath22 g , while the field strength in the centre of the flux rope reaches 10 g. together with the isothermal @xmath23 mk assumption and the plasma densities of @xmath24 @xmath2 and @xmath25 @xmath2 at the bottom boundary and in the flux rope , respectively , this implies that free - free emission and absorption is the only relevant mechanism for the radio emission . therefore the synthetic radio views in this case
reflect only the plasma density structure from the mhd data .
the isothermal mhd evolution self - consistently incorporates density changes and ultimately results in a delicate force - balance between lorentz force , pressure gradient and gravity throughout the flux rope structure .
this flux rope represents the magnetic structure of the filament channel of a quiescent filament . by taking snapshots corresponding to the times @xmath26 , 53 and 80 ( with the time unit chosen to be 86 seconds throughout this paper ) , we then obtain successive radio views on the forming flux ropes where the frames are about 38 minutes apart in real time ; these radio views are presented in section [ s - quiet ] .
besides the low - field case from @xcite targeting quiescent filaments , we also analyse two more cases that represent smaller flux rope configurations at increased field strengths to mimic the magnetic structure of active region filaments .
a medium field strength case has been generated by a procedure analogous to the low - field one , with magnetic field strength at the photosphere level reaching @xmath27 g , ultimately attaining about @xmath28 g in the flux rope centre . in this case , the bottom plane of the simulation box is located at 4 mm above the photosphere representing the low corona where the field strength then reaches @xmath29 g ; the ( constant ) temperature and density are similar to the low - field model .
finally , a high field strength case has been computed as well . in this case , the magnetic field strength reaches @xmath30 g at the photosphere level , @xmath31 g at the bottom plane of the simulation box ( in the low corona ) , and about 680 g in the flux rope ; the density and temperature conditions are set to @xmath32 @xmath2 ( at the bottom plane ) and @xmath33 mk , respectively .
as said above , the simulations setup in the medium - field and high - field cases is similar to that in the study of @xcite , but the computational box size is reduced to 120 mm ( length ) , 100 mm ( width ) , and 60 mm ( height ) , and the spatial resolution is reduced to 1 mm .
the reason is that the mhd simulations for an increased magnetic field strength ( and the same spatial resolution ) would require much shorter time steps to satisfy the stability criterion @xcite . due to the limited computational resources
, we choose to reduce the resolution of the stronger field cases , although this increases the numerical resistivity and reduces the spatial accuracy .
ultimately , the main difference among the three models is in the attained maximum field strength , while the magnetic topologies ( important for later filament formation ) are very similar , despite of the differences in the box size , resolution , and thermodynamic conditions .
although the flux rope formation timescales differ in the considered three cases , here we are mainly concerned with the endstates obtained .
in contrast to the low - field case @xcite , the endstate in the medium- and high - field models is defined more qualitatively : as a state when the overall magnetic structure does not change anymore .
figure [ figchunmh ] gives an impression of both the medium- and high - field strength magnetic topologies and the corresponding plasma density distributions in their final stable flux rope endstates . in the adopted numerical time units these endstates
are reached at @xmath34 and @xmath35 for the medium- and high - field cases , respectively .
the mhd simulation results are used below to produce typical radio views for the flux ropes located either on the solar disk or at the limb ; the synthetic radio views are presented in sections [ armf ] and [ arhf ] for the medium- and high - field models , respectively .
we should note that our models are based on the full set of mhd equations and thus go beyond the oft - used force - free approximation for the coronal magnetic field .
this allows formation of strong magnetic fields in the corona ; in particular , in the high - field model of an active region the magnetic field in the flux rope can become stronger than at the bottom of the simulation box . in turn
, strong magnetic fields can result in formation of coronal gyroresonance emission sources .
the final synthetic radio maps we present are based on the endstate of the fully thermodynamically consistent mhd study @xcite . in this case the low - field isothermal model was used as an input to study how optically thin radiative losses could trigger a full prominence formation within the dipped magnetic field region of the flux rope .
a stable prominence structure with overall dimensions of @xmath36 mm was established with the average density of @xmath37 @xmath38 , i.e. , well above the surrounding coronal density of @xmath39 @xmath38 .
although the structure is relatively small and low in mass , its temperature contrast with the corona is realistic with a @xmath40 k filament embedded in a 1 mk corona .
the magnetic topology and plasma distribution corresponding to the final state of the filament evolution are shown in figure [ filamentxc ] . in this work , this low - field model with embedded prominence
is used to generate the first synthetic views on simulated filaments ( see section [ res - filament ] ) .
as has been said above , the dominant mechanism determining the appearance of the solar filaments at radio wavelengths is the free - free emission and absorption .
this process can be described by the radiation transfer equation @xmath41 where @xmath42 is the spectral emission intensity of the wave mode @xmath43 ( either ordinary or extraordinary ) and @xmath44 is the distance along the line of sight .
the spectral emissivity @xmath45 and the absorption coefficient @xmath46 for the thermal bremsstrahlung process are given by the equations ( * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
@xmath47 where @xmath48 is the emission frequency , @xmath49 is the refraction index ( under the considered conditions , it is close to unity for both wave modes ) , @xmath7 is the plasma temperature , @xmath50 is the electron concentration , the factor @xmath51 describes the average ion charge and @xmath52 is the coulomb logarithm . in the solar corona
, the ion charge factor can be taken to be @xmath53 at the temperatures above @xmath54 k ( the helium ionization temperature ) or @xmath55 below that temperature .
the coulomb logarithm ( in the cgs units ) equals approximately @xmath56 note that the emission does not propagate ( its absorption coefficient @xmath46 goes to infinity ) at the frequencies below the plasma frequency .
another mechanism that can be responsible for the radio emission of solar filaments and filament channels is the gyroresonance emission of the thermal electrons gyrating in magnetic field ( e.g. , * ? ? ?
* ; * ? ? ?
in this case , the emission ( and/or absorption ) takes place at narrow resonance layers where the emission frequency coincides with a harmonic of the local electron cyclotron frequency @xmath57 ( i.e. , @xmath58 , @xmath59 ) .
the change of the emission intensity at a given gyrolayer can be described by the equation @xmath60,\ ] ] where @xmath61 and @xmath62 are the emission intensities before and after the gyrolayer , and @xmath63 is the gyrolayer s optical depth : @xmath64 in the above equation , @xmath65 is the viewing angle ( the angle between the local magnetic field and the line of sight ) , @xmath66 and @xmath67 are the components of the wave polarization vector ( see , e.g. , * ? ? ? * ) , and @xmath68 is the inhomogeneity scale of the magnetic field in the direction along the line of sight .
evidently , the gyroresonance emission requires a sufficiently strong magnetic field ; in our simulations , it has been found to be important only in the high - field active region model ( see section [ arhf ] ) . under the conditions typical of the solar corona , the gyrolayers corresponding to the harmonics with @xmath69 or
@xmath70 are expected to be optically thick ( @xmath71 ) or optically thin ( @xmath72 ) , respectively . as a result , the observed emission ( with the brightness temperature equal to the local plasma temperature ) is expected to be produced at the third gyrolayer , although under certain conditions other cyclotron harmonics can become dominant .
we calculate the two - dimensional emission maps by numerical integration of the radiation transfer equations ( [ transfer ] , [ gr ] ) along the selected lines of sight ( corresponding to the image pixels ) .
the initial intensity ( at the far boundary of the simulation box ) is set either to zero ( for the filaments and flux ropes located at the limb ) or to the typical emission intensity of the quiet sun ( for the sources on the solar disk ) ; the average radio spectrum of the quiet sun is taken from the paper of @xcite .
the radio emission is calculated using the numerical code developed by @xcite , which takes into account the contributions of both the free - free and gyroresonance emission mechanisms .
in addition to the above mentioned factors , this code automatically calculates the local temperature - dependent ionization degree ( which affects the electron concentration @xmath50 ) using the saha equation @xcite ; for the typical densities in the solar corona ( @xmath73 @xmath2 ) , the ionization degree rises steeply from nearly zero to nearly 100% at the temperature of about @xmath74 k.
as follows from observations and is supported by theory , the appearance of the filaments and filament channels in radio emission is frequency - dependent . on the other hand ,
the angular resolution of the radio telescopes is also frequency - dependent ; in particular , in the decimetric / metric range both the existing instruments ( e.g. , the nanay radioheliograph ) and the instruments expected soon will be unable to resolve any source structure for the considered mhd models .
therefore we consider here only the microwave emission ( at the frequencies of about @xmath75 ghz ) .
this frequency range is covered by such imaging instruments as the siberian solar radio telescope ( 5.7 ghz ) and nobeyama radioheliograph ( 17 and 34 ghz ) ; several instruments are currently under construction , such as the expanded owens valley solar array ( @xmath76 ghz ) , upgraded siberian solar radio telescope ( @xmath77 ghz ) and chinese spectral radioheliograph ( @xmath78 ghz ) .
( black ) to @xmath79 ( white ) ; the corresponding range of brightness temperatures @xmath80 $ ] is given at the right side of each panel . ] figure [ quietevol ] demonstrates the simulation results for the quiet - sun filament channel model . in this model ,
the magnetic field is weak ( less than 20 g ) , the plasma density varies from @xmath81 to @xmath82 @xmath2 ; the plasma temperature is set to 1 mk .
three time moments illustrating different stages of the flux rope evolution are shown , in the simulation they are about 38 minutes apart .
we have found that the free - free emission at the frequencies above 1 ghz is optically thin see the spectra in figure [ spectrafinalstate ] ) ; the contribution of the gyroresonance emission is negligible .
therefore the computed images at all considered frequencies are nearly the same ( except of different values of the brightness temperature ) and coincide with the euv images as well . in figure [ quietevol ] , the frequency of 17 ghz ( one of the working frequencies of the nobeyama radioheliograph ) is chosen . if the filament channel is observed from above ( i.e. , located at the center of the solar disk ) then at the beginning of the evolution process we see a bright stripe aligned along the photospheric neutral line and two darker areas to right and left from it .
later , the strongest emission gradually concentrates in two bright spots located at some distance from the neutral line ; note an increase of the maximum radio brightness with time .
however , the emission intensity is still relatively low : the relative radio brightening ( after subtraction of the quiet sun emission level ) does not exceed 7500 k at 17 ghz and 90000 k at 5 ghz .
although such brightenings can be detectable , they are too weak in comparison with the observed neutral - line - associated sources in active regions .
if the source region is located at the solar limb then the images are dominated by two bright spots at the footpoints of the flux rope .
the coronal emission is much weaker , which makes it hard to detect .
nevertheless , one can notice a faint arc - like structure that slowly ascends with time .
[ [ armf ] ] medium - field model + + + + + + + + + + + + + + + + + + figure [ armfevol ] demonstrates the evolution of an active - region filament channel
. the magnetic field strength in the considered model can reach 100 g , which exceeds the typical values for the quiet sun , but is still relatively weak for the solar active regions .
the plasma density varies from @xmath83 to @xmath84 @xmath2 ; the plasma temperature is set to 1 mk .
like in the previous case , the free - free emission at the frequencies above 1 ghz is optically thin and the contribution of the gyroresonance emission is negligible ; therefore we show only the 17 ghz images .
the frames are taken at times @xmath85 , 66 and 100 , i.e. , they also cover the deformation phase before the flux rope formation ( the flux rope reaches a stable configuration at @xmath34 , as shown in figure [ figchunmh ] , top panel ) .
although the simulation box is now smaller than in the previous model , the source structure and evolution are qualitatively similar .
initially , the emission is produced mainly in a narrow stripe along the photospheric neutral line ( when seen from above ) , and later concentrates in two bright spots corresponding to the footpoints of the forming flux rope .
the brightness temperatures are also similar to the previous model , i.e. , much lower than in the observed neutral - line - associated microwave sources .
the limb view demonstrates a gradually rising arc - like structure .
[ [ arhf ] ] high - field model + + + + + + + + + + + + + + + + figures [ arhfevol17][arhffinal3 ] correspond to the high - field model of the active - region filament channel ( figure [ figchunmh ] , bottom panel ) .
now , the magnetic field is stronger ( up to 1000 g at the solar surface and up to 680 g within the flux rope ) .
the plasma density varies from @xmath86 to @xmath87 @xmath2 ; the plasma temperature is set to 1.5 mk .
figure [ arhfevol17 ] demonstrates the evolution of the microwave emission at the frequency of 17 ghz , where the contribution of the gyroresonance emission is negligible ( i.e. , only the free - free emission is present ) . like in the previous cases ,
the microwave images reflect the spatial distribution of the plasma within the simulated active region .
the top view also reveals some similarity to the models with lower magnetic field : initially , the emission is concentrated in a short bright stripe aligned approximately along the photospheric neutral line ; later , this stripe turns slowly with respect to the photospheric neutral line , and its ends gradually become brighter . in the limb view , we can see that the emission is produced mainly at some height above the solar surface ; the evolution of the microwave images reflects the formation of an overdense region in the corona .
as the plasma is accumulated within the flux rope , the emission brightness temperature gradually increases .
this increase is much faster than in the previous ( low - field and medium - field ) cases , and the resulting brightness temperature is much higher as well : it can exceed 0.15 mk .
evidently , this is caused by the higher concentration of plasma within the flux rope ( up to @xmath87 @xmath2 ) . as a result ,
the free - free emission ( although still optically thin , with @xmath88 ) becomes much brighter .
our simulations indicate that , at least , some of the observed neutral - line associated sources with moderate brightness temperatures ( see , e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) can be explained by the purely thermal free - free emission from overdense flux ropes .
the simulated images at higher frequencies ( above 17 ghz ) are not shown because they are nearly the same as in figure [ arhfevol17 ] , except of different values of the brightness temperature ; this behaviour is typical of the optically thin free - free emission .
, for the emission frequency of 5 ghz . ]
the contribution of the gyroresonance emission becomes noticeable ( or even dominant ) at the frequencies below @xmath89 ghz ( see the spectra in figure [ spectrafinalstate ] ) , which affects the microwave images dramatically .
figure [ arhfevol5 ] demonstrates the evolution of the microwave emission at the frequency of 5 ghz . initially ( at @xmath90 and 10 ) , when the magnetic field in the core region of the forming flux rope is relatively weak , the emission at the considered frequency is produced mainly at the fourth gyrolayer .
this gyrolayer is optically thin , i.e.
, the emission brightness temperature is lower than the plasma temperature ; nevertheless , the contribution of the gyroresonance emission far exceeds that of the free - free emission . as a result ,
the microwave images reflect mainly the structure of the magnetic field in the active region .
as the flux rope grows and its magnetic field strength increases , the third gyrolayer becomes the dominant emission source at the considered frequency ; this gyrolayer is optically thick , i.e. , the emission brightness temperature approaches the plasma temperature of 1.5 mk . at the final stage of the filament channel evolution ( at @xmath35 ) ,
the brightest ( saturated ) areas in the simulated microwave images represent the projections of the third gyrolayer , i.e. , of the corresponding magnetic field strength isosurface ( @xmath91 g for the emission frequency of 5 ghz ) .
note that , as can be seen in the limb view at @xmath35 , this isosurface is located at some height in the corona and separated from the bottom of the simulation box .
such geometry reflects the above - mentioned fact that the considered model is not force - free ; as a result , concentration of the magnetic field within the flux rope can create field strengths higher than in the lower layers .
this effect enables generation of the neutral - line - associated sources due to the gyroresonance emission mechanism even in the active regions where the photospheric magnetic fields seem to be insufficient for that ( cf .
* ; * ? ? ?
[arhfevol5 ] , at @xmath35 ) at the frequency of 3 ghz . left : the source region is located at the center of the solar disk ; right : the source region is located at the limb . ]
figure [ arhffinal3 ] shows the simulated radio images at the frequency of 3 ghz ( only the final state of the simulation at @xmath35 is shown ) .
again , the emission is produced mainly at the third gyrolayer .
the brightest ( saturated ) areas in the images reflect the shape of this gyrolayer that corresponds now to the @xmath92 g isosurface ; the brightness temperature reaches the plasma temperature of 1.5 mk .
the above simulations demonstrate that the gyroresonance emission from a hot dense flux rope with a strong magnetic field can be responsible , at least , for some of the observed neutral - line - associated microwave sources in the solar active regions .
the brightness temperature of the emission in our simulations reaches 1.5 mk , which equals the assumed plasma temperature in the considered model .
such brightness temperatures are quite common for the observed neutral - line - associated sources , as reported , e.g. , by @xcite . on the other hand ,
some observations have reported considerably higher brightness temperatures of the neutral - line - associated sources : up to @xmath93 mk ( * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* etc . ) .
evidently , such sources would require a higher plasma temperature , which seems to be feasible in the active regions @xcite .
as said above , the magnetic field strength in the considered model ( in the flux rope ) can reach 680 g. therefore the gyroresonance emission mechanism is limited by the frequencies of up to @xmath94 ghz ( for the emission at the third - fourth cyclotron harmonic ) . to produce emission at higher frequencies , we need a stronger magnetic field .
for example , the emission at 17 ghz requires the field strength of @xmath95 g ( again , for the emission at the third - fourth cyclotron harmonic ) , which is quite typical of the solar active regions .
@xcite reported the formation of a simulated filament by in situ condensation in a pre - existing flux rope using thermodynamic mhd simulations .
the final stable state is shown in figure [ filamentxc ] .
we have calculated the radio emission from this model filament , and figures [ filamentfinal][filamentfinalsm ] demonstrate the simulated radio images at three frequencies ; the final stage of the filament evolution is shown .
the plasma density in the filament reaches @xmath96 @xmath2 and the temperature can drop to about @xmath97 k. the magnetic field in the simulation box does not exceed 11 g ; therefore the contribution of the gyroresonance emission is negligible .
the chosen frequencies correspond to the working frequencies of the siberian solar radio telescope ( 5.7 ghz ) and the nobeyama radioheliograph ( 17 and 34 ghz ) . )
convolved with the instrumental response functions at different frequencies . at the right side of each panel ,
the range of brightness temperatures in the image is given . ]
figure [ filamentfinal ] shows the non - smoothed images .
in addition , to account for a limited resolution of the existing radio telescopes , the calculated emission maps were convolved with the model ( symmetric gaussian - shaped ) instrumental response functions . for the filaments at the limb , the image smoothing procedure takes into account the contribution of the solar disk ( with the typical intensity of the quiet sun ) at the corresponding ( left ) boundary of the image .
the beam widths are taken to be 21 , 10 and 5 at the frequencies of 5.7 , 17 and 34 ghz , respectively .
the resulting ( smoothed ) radio maps are shown in figure [ filamentfinalsm ] .
since the model filament is hotter than real ones ( @xmath98 k ) , the filament on the disk looks brighter than the underlying chromosphere .
in addition , the estimation of @xmath99 k refers only to the central part of the filament , while the plasma in the transition region between the filament and the surrounding corona is much hotter and still has a sufficiently high density . as a result , this transition region makes a dominant contribution to the radio emission , which is seen especially well in the non - smoothed image at 5.7 ghz in figure [ filamentfinal ] ; the brightness temperature reaches 49000 k. smoothing due to the limited spatial resolution removes this fine filament structure and lowers the image contrast ( see figure [ filamentfinalsm ] ) . in general
, the simulated radio images at the disk center look qualitatively different from the observed ones due to the higher filament temperature . nevertheless , if the filament is observed against a bright active region ( like , e.g. , in the event reported by * ? ? ?
* ) , it will appear as a darker stripe .
in contrast , the filament above the limb looks qualitatively similar to the observed ones , i.e. , like a bright prominence ( although the brightness temperatures are higher than in real observations ) . at the frequencies of @xmath100 ghz ,
the filament is optically thick and its brightness temperature ( in the non - smoothed images ) reflects an effective temperature of the transition region between the filament and corona ( @xmath101 k at 5.7 ghz ) . at higher frequencies ,
the filament becomes optically thin . from the non - smoothed images at 17 and 34
ghz we can conclude that its brightness temperature decreases with frequency @xmath48 approximately as @xmath102 ; this behaviour is typical of the optically thin free - free emission and has been detected in the multi - frequency observations of solar filaments .
the visible structure of the filament at 17 and 34 ghz ( again , in non - smoothed images ) reflects the plasma and temperature distribution .
since the considered model filament is relatively small , convolving the simulated images with the instrument response functions removes any fine structure ( see figure [ filamentfinalsm ] ) ; at lower frequencies , the filament merges with the solar disk .
as has been said above , the parameters of our mhd models have been chosen to illustrate different regimes of the flux rope formation . besides the different magnetic field strengths , these models can have ( due to the numerical restrictions ) different simulation box sizes and different initial plasma parameters
. therefore a direct quantitative comparison of the models is hardly possible .
nevertheless , we summarize here some representative quantities in order to provide a more comprehensive picture of our simulations and to analyze general trends
. corresponds to an unperturbed unmagnetized corona .
a ) the quiet - sun filament channel ( low - field model ) ; a@xmath103 ) the simulated filament ( no time dependence ) ; b ) the medium - field active - region filament channel model ; c ) the high - field active - region filament channel model .
solid lines : emission from a flux rope footpoint ; dashed lines : emission from the flux rope center . in panel ( a@xmath103 ) , the asterisks ( @xmath104 ) show the emission from the center of the simulated filament in its final stable state.,title="fig : " ] corresponds to an unperturbed unmagnetized corona .
a ) the quiet - sun filament channel ( low - field model ) ; a@xmath103 ) the simulated filament ( no time dependence ) ; b ) the medium - field active - region filament channel model ; c ) the high - field active - region filament channel model . solid lines : emission from a flux rope footpoint ; dashed lines : emission from the flux rope center . in panel ( a@xmath103 )
, the asterisks ( @xmath104 ) show the emission from the center of the simulated filament in its final stable state.,title="fig : " ] corresponds to an unperturbed unmagnetized corona .
a ) the quiet - sun filament channel ( low - field model ) ; a@xmath103 ) the simulated filament ( no time dependence ) ; b ) the medium - field active - region filament channel model ; c ) the high - field active - region filament channel model .
solid lines : emission from a flux rope footpoint ; dashed lines : emission from the flux rope center . in panel ( a@xmath103 ) , the asterisks ( @xmath104 ) show the emission from the center of the simulated filament in its final stable state . ] corresponds to an unperturbed unmagnetized corona .
a ) the quiet - sun filament channel ( low - field model ) ; a@xmath103 ) the simulated filament ( no time dependence ) ; b ) the medium - field active - region filament channel model ; c ) the high - field active - region filament channel model .
solid lines : emission from a flux rope footpoint ; dashed lines : emission from the flux rope center . in panel ( a@xmath103 ) , the asterisks ( @xmath104 ) show the emission from the center of the simulated filament in its final stable state . ] figure [ lightcurves ] demonstrates the temporal evolution of the flux ropes .
namely , we plot here the microwave light curves at two frequencies ( 5 and 17 ghz ) for two chosen locations ( pixels ) in the image plane ; the flux ropes are assumed to be located at the center of the solar disk .
the flux rope center corresponds to the coordinates @xmath105 , @xmath106 in figures [ quietevol][filamentfinalsm ] ( top views ) , while the tentative flux rope footpoint location at each time is defined as the point of the brightest optically thin radio emission .
the light curves for three isothermal models are shown ; in the top row of figure [ lightcurves ] we show also the corresponding brightness temperatures for the non - isothermal filament model ( the endstate and the central pixel only ) .
the background ( initial ) emission is subtracted , i.e. , the values in figure [ lightcurves ] represent the relative microwave brightening caused by the plasma and magnetic field concentration .
evidently , different models have different evolution timescales : in the models with stronger magnetic field , the flux ropes are developing faster .
as has been said above , the microwave emission in the low- and medium - field models is produced due to the free - free mechanism in the optically thin regime ; as a result , the 5 and 17 ghz light curves mimic each other ( they differ by a constant factor ) .
the emissions from the flux rope center and from the footpoints demonstrate different dynamics ; at later stages of the evolution , the emission from the footpoints region becomes considerably brighter due to larger projected column density ( for the considered orientation ) .
in contrast , in the high - field model at low frequencies the gyroresonance emission mechanism becomes increasingly dominant with time ; at later stages of the evolution , the emission brightness temperature at @xmath107 ghz approaches the plasma temperature ( both at the flux rope center and at the footpoints ) .
the intensity of the optically thin free - free emission ( e.g. , at 17 ghz ) in this model steadily increases with time , with a relatively low contrast between the flux rope center and the footpoints .
finally , the emision brightness temperatures for the filament core are higher than the corresponding values for the isothermal low - field flux rope , due to both the higher plasma density and the lower temperature in the filament .
figure [ spectrafinalstate ] demonstrates several examples of broadband microwave spectra of the simulated flux ropes and the filament ( in their endstates ) . like in the previous figure ,
the flux ropes are assumed to be located at the center of the solar disk ; the emissions from the flux rope center and footpoint are considered .
the typical spectrum of the quiet sun @xcite is plotted for reference .
we can see that the spectra for the low- and high - field models almost coincide ; they look as the typical isothermal free - free spectra ( although the plasma densities can be somewhat different in different models and parts of the flux rope ) . in the high - field model ,
the spectra include the optically thick gyroresonance part corresponding to the third gyrolayer ( at the frequencies of up to @xmath108 ghz ) , the optically thin gyroresonance part corresponding to the fourth gyrolayer ( up to @xmath89 ghz ) and the optically thin free - free part ( at higher frequencies ) ; such spectral structure can be potentially used for diagnosing the coronal magnetic field , provided that we have observations with sufficiently high spectral and spatial resolutions .
finally , the filament spectrum reflects the dominant emission mechanism in that model , i.e. , the free - free emission in a multi - thermal source .
we have demonstrated the ability to generate virtual radio views on solar flux ropes from mhd simulations , where full thermodynamic information is available .
they can be contrasted with the actual radio views at different frequencies , and may serve to verify inversions ( i.e. , estimates of the field strengths and density - temperature conditions ) made from the radio data .
we have shown how the views for different orientations ( either on the disk or at the limb ) can be produced , with ( possibly ) taking into account the instrument - specific viewing conditions .
the flux ropes with low to medium magnetic field strengths produce radio emission due to the thermal bremsstrahlung process ; a deformation from an arcade to a flux rope configuration is seen as a clear change from a photospheric - inversion - line - aligned emission feature to a footpoint - dominated double - spot pattern . at higher field strengths , the gyroresonance emission mechanism dominates at @xmath109 ghz ;
the resulting radio views can be associated with the frequently observed neutral - line - associated microwave sources .
the medium - field model representing the intermediate case ( with the field strength of the order of 100 g ) provides the results qualitatively similar to the low - field model ; nevertheless , it is an important illustration of the transition from the low - field to high - field mhd models
. for the virtual filament studied , our on - disk images show emission instead of the expected brightness depression . in the limb views ,
the filament can clearly be seen above the limb , especially at higher frequencies ( @xmath110 ghz ) .
we conclude that the approach to simulating the filaments seems correct , although the simulation parameters ( i.e. , the filament temperature and dimensions ) still differ from the typical parameters of the observed filaments .
since only a few mhd models have been analyzed ( because , as said above , the mhd simulations are very expensive computationally ) , our results are by no means comprehensive . instead
, this study should be considered as a first attempt to combine the mhd simulations of the flux ropes and filaments with the 3d simulations of the radio emission .
the results obtained , although somewhat qualitative , demonstrate how the flux rope evolution affects the radio emission ; they show the directions for further , more advanced and realistic simulations .
this work was supported in part by the russian foundation of basic research ( grants 14 - 02 - 91157 , 15 - 02 - 03717 and 15 - 32 - 51171 ) and by a marie curie international research staff exchange scheme `` radiosun '' ( people-2011-irses-295272 ) .
rk and cx acknowledge support from the project goa/2015/014 ( ku leuven ) , the interuniversity attraction poles programme initiated by the belgian science policy office ( iap p7/08 charm ) , and the fwo .
part of the simulations used the vsc ( flemish supercomputer center ) funded by the hercules foundation and the flemish government .
81 # 1isbn # 1#1#1#1#1#1#1#1_#1_#1*#1*#1#1#1#1#1#1#1#1#1#1#2#1#1#1#1#1_#1_#1#1#1*#1*#1#1#1_#1_#1#1#1#1#1#1#1#1#1#1#1 # 1http://dx.doi.org/#1 [ ] # 1http://arxiv.org / abs/#1 [ ] # 1http://adsabs.harvard.edu / abs/#1[]#1#1#1#1#1#1#1#1#1#1#1#1 | in this paper , we produce synthetic radio views on simulated flux ropes in the solar corona , where finite - beta magnetohydrodynamic ( mhd ) simulations serve to mimic the flux rope formation stages , as well as their stable endstates .
these endstates represent twisted flux ropes where balancing lorentz forces , gravity and pressure gradients determine the full thermodynamic variation throughout the flux rope .
the obtained models are needed to quantify radiative transfer in radio bands , and allow us to contrast weak to strong magnetic field conditions .
field strengths of up to 100 g in the flux rope yield the radio views dominated by optically thin free - free emission .
the forming flux rope shows clear morphological changes in its emission structure as it deforms from an arcade to a flux rope , both on disk and at the limb . for an active region filament channel with a field strength of up to 680 g in the flux rope , gyroresonance emission ( from the third - fourth gyrolayers )
can be detected and even dominates over free - free emission at the frequencies of up to 7 ghz .
finally , we also show synthetic views on a simulated filament embedded within a ( weak - field ) flux rope , resulting from an energetically consistent mhd simulation . for this filament ,
synthetic views at the limb show clear similarities with actual observations , and the transition from optically thick ( below 10 ghz ) to optically thin emission can be reproduced . on the disk ,
its dimension and temperature conditions are as yet not realistic enough to yield the observed radio brightness depressions . |
following milo et al .
@xcite , we perform 3- and 4-node motif detection on the transcriptional regulatory networks of _ e. coli _ ( version @xmath0 ) and _ s. cerevisiae _ , using their freely available network data and software ( mfinder version @xmath1 ) @xcite . generation of randomized networks from the actual network is performed according to one of three null models : an edge - swapping algorithm , an edge - matching algorithm , and a monte carlo algorithm , all described in detail in @xcite . because significance results are similar among models ( cf .
_ results _ and @xcite ) , emphasis in this note is placed on the edge - swapping algorithm , a markov chain procedure that repeatedly swaps the target nodes between pairs of edges .
@xmath2-scores are computed from the mean and standard deviation of the count of a particular subgraph within an ensemble of at least 1,000 randomized networks @xcite .
we quantify correlation between the counts of any two subgraphs over the course of a randomization process using mutual information @xcite .
mutual information captures correlation between two random variables even when a relationship exists that is nonlinear ( unlike , e.g. , the correlation coefficient ) or non - monotonic ( unlike , e.g. , spearman s rho ) . in this study , the counts @xmath3 and @xmath4 of the @xmath5th and @xmath6th subgraphs at each iteration of the edge - swapping process are used to increment a counts matrix from which the joint probability distribution @xmath7 is obtained by normalization . mutual information @xmath8 is computed as @xmath9 where the log is base 2 to give @xmath8 in bits , and @xmath10 and @xmath11 .
mutual information is bounded from below by @xmath12 ( as seen in eqn .
[ i ] when there is no correlation and the subgraph counts are independent of each other i.e. @xmath13 ) and bounded from above by the smaller of the two variables entropies , where @xmath14 and the analogous expression with @xmath15 are the entropies of the @xmath5th and @xmath6th subgraphs counts respectively . in order to obtain a statistic that can be compared across all subgraph pairs ,
we normalize by the average entropy , defining @xmath16 as our measure of correlation .
note that @xmath17 , with @xmath18 when @xmath19 and @xmath20 when @xmath21 .
we find qualitatively similar results ( cf .
_ results _ ) when normalizing by the minimum , instead of the average , entropy .
only four 3-node subgraphs are present in the transcriptional network of _
e. coli _ , and a @xmath2-score analysis of the type performed in @xcite reveals a curious effect .
specifically , with respect to ensembles generated via any of the edge - swapping , edge - matching , and monte carlo algorithms @xcite , the @xmath2-scores of three of the subgraphs ( ids 6 , 12 , and 36 ; cf .
[ dict ] ) are either very close or equal to the negative of the @xmath2-score of the fourth subgraph ( the feed - forward loop , i d 38 ) ; see fig .
[ table ] .
in fact , as shown in fig . [ overlap ] , the absolute value of the difference in counts within the actual network and counts within a sample randomized network at each iteration of the edge - swapping algorithm is the same among all four subgraphs for the first 1,000 iterations .
the interpretation is simple : as detailed in fig .
[ toy ] , each time an edge of a feed - forward loop is swapped with an external edge , the feed - forward loop is destroyed and one of each of the other three subgraphs is created ; using subgraph ids we may denote this process as @xmath22 since this process accounts for the overwhelming majority of the changes in count of the latter three subgraphs , there is extremely high correlation among the counts of all four subgraphs in each randomized network , and the magnitudes of their @xmath2-scores are very close . -scores are calculated with respect to an ensemble of 1,000 randomized networks , generated via the edge - swapping algorithm @xcite .
plots show the count of each subgraph during the generation of one randomized network .
each iteration corresponds to one edge - swap . ] in the actual network and count @xmath23 at each iteration of the edge - swapping algorithm , for the subgraphs in fig .
[ table ] .
note that all four curves completely overlap . ] to quantify and extend the detection of correlations such as that just described , we use a normalized mutual information measure , as detailed in _
methods_. for the cases of 3- and 4-node subgraphs in both the _ e. coli _ and _ s. cerevisiae _ transcriptional networks , the measure @xmath24 ( cf .
[ a ] ) is computed between all pairs of subgraphs @xmath5 and @xmath6 that appear during the randomization of a network via the edge - swapping algorithm .
[ mi3]-[mi4 ] show the matrices @xmath24 ; the row and column order is determined by summing along either direction and sorting , which tends to group together sets of subgraphs with high pairwise correlations . during randomization of the _ e. coli _ network , a set of four 3-node subgraphs ( ids 6 , 12 , 36 , and 38 ; cf .
[ dict ] ) are highly correlated , as shown by the bright 4-by-4 square in fig .
the high correlation is simply the result of the effect described in the previous section , in which any of three swaps overwhelmingly converts a feed - forward loop ( i d 38 ) into three other subgraphs ( ids 6 , 12 , and 36 ) .
in fact the same set of high correlations in seen during the randomization of the _ s. cerevisiae _ network , as shown by the upper left 4-by-4 square in fig .
there are additional correlated sets in _ s. cerevisiae _ :
subgraphs 14 , 74 , and 102 are highly correlated as indicated by the bright 3-by-3 square involving these ids in fig .
[ mi3]b , and subgraphs 74 and 108 , as well as 14 and 46 , are correlated as indicated by the relatively bright entries at these coordinate pairs in fig .
[ mi3]b . respectively ,
these correlations are due to the effects ( in the notation of eqn .
[ eff1 ] ) @xmath25 of which one may convince oneself with the aid of fig .
note that although subgraphs 14 , 102 and 108 participate in the highly correlated effects described here , none changes in number significantly enough upon randomization to be labeled a motif in the _ s. cerevisiae _ network ( subgraphs 46 and 74 do not appear in the actual network , only during the course of the randomization ) .
our analysis reveals correlations between counts of 4-node subgraphs as well . as indicated by the bright blocks and off - diagonal elements in fig .
[ mi4 ] , several sets of subgraphs are highly correlated during the randomization of both the _ e. coli _ and _ s. cerevisiae _ networks .
correlations are less easily interpreted in the 4-node case than in the 3-node case , but one must nonetheless remain aware of such artifacts of the randomization process when identifying subgraphs as motifs .
we note that the bi - fan ( i d 204 ) , the 4-node subgraph commonly identified as a motif in a variety of networks including both transcriptional networks studied here @xcite , does not exhibit particularly high correlation with any other subgraph under our measure in either the _ e. coli _ or _ s. cerevisiae _ network .
we find results qualitatively similar to figs .
[ mi3]-[mi4 ] when normalizing by the minimum , instead of the average , entropy in eqn .
the technique we describe here can be extended to the detection of subgraphs of any size .
[ a ] ) between all pairs of 3-node subgraphs @xmath5 and @xmath6 that appear during the randomization of a network via the edge - swapping algorithm @xcite for the transcriptional networks of _ e. coli _ ( a ) and _ s. cerevisiae _ ( b ) .
subgraphs are labeled as in fig .
the row and column order is determined by summing along either direction and sorting.,title="fig : " ] ( cf .
[ a ] ) between all pairs of 3-node subgraphs @xmath5 and @xmath6 that appear during the randomization of a network via the edge - swapping algorithm @xcite for the transcriptional networks of _ e. coli _ ( a ) and _ s. cerevisiae _ ( b ) .
subgraphs are labeled as in fig .
the row and column order is determined by summing along either direction and sorting.,title="fig : " ] ( cf .
[ a ] ) between all pairs of 4-node subgraphs @xmath5 and @xmath6 that appear during the randomization of a network via the edge - swapping algorithm @xcite for the transcriptional networks of _ e. coli _ ( a ) and _ s. cerevisiae _ ( b ) .
subgraphs are labeled as in alon et al.s `` motif dictionary '' @xcite .
the row and column order is determined by summing along either direction and sorting.,title="fig : " ] ( cf .
[ a ] ) between all pairs of 4-node subgraphs @xmath5 and @xmath6 that appear during the randomization of a network via the edge - swapping algorithm @xcite for the transcriptional networks of _ e. coli _ ( a ) and _ s. cerevisiae _ ( b ) .
subgraphs are labeled as in alon et al.s `` motif dictionary '' @xcite .
the row and column order is determined by summing along either direction and sorting.,title="fig : " ]
by quantifying correlations among subgraph counts during 3- and 4-node motif detection in the transcriptional networks of _ e. coli _ and _ s. cerevisiae _ , we reveal that motifs come in sets : the destruction of a subgraph during the randomization process can be highly correlated with the creation of one or more other subgraphs .
the correlations are easily understood in the 3-node case , and we present an information - theoretic tool to extract such correlations in general .
it has not escaped our attention that this observation serves as the basis for a more principled clustering of subgraphs based on correlations ( e.g. , by mixture - modeling in which the state of the subgraph count is a mixture of several states , with counts conditionally independent given the state ) .
the correlations among subgraphs are artifacts of the algorithm used to generate the ensemble of randomized networks ; although we demonstrate their existence here in the context of only one randomization algorithm , the edge - swapping algorithm , they occur in other commonly used algorithms , as evidenced by mutually consistent effects on the @xmath2-scores .
these findings do not necessarily invalidate the statuses of commonly identified motifs ( it remains the case , for example , that there are significantly more feed - forward loops in the transcriptional network of _
e. coli _ than in a random network generated under most any commonly used null model ) ; they do argue , however , that the limitations of the randomization scheme should be fully recognized during the motif finding process . | the identification of motifs subgraphs that appear significantly more often in a particular network than in an ensemble of randomized networks has become a ubiquitous method for uncovering potentially important subunits within networks drawn from a wide variety of fields .
we find that the most common algorithms used to generate the ensemble from the real network change subgraph counts in a highly correlated manner , so that one subgraph s status as a motif may not be independent from the statuses of the other subgraphs .
we demonstrate this effect for the problem of 3- and 4-node motif identification in the transcriptional regulatory networks of _ e. coli _ and _ s. cerevisiae _ in which randomized networks are generated via an edge - swapping algorithm ( milo et al . ,
_ science _ * 298*:824 , 2002 ) .
we show that correlations among 3-node subgraphs are easily interpreted , and we present an information - theoretic tool that may be used to identify correlations among subgraphs of any size .
identifying motifs has become a standard way to probe the functional significance of biological , technological , and sociological networks @xcite .
a motif is commonly defined as a subgraph whose number of appearances in a particular network is significantly greater than its average number of appearances in an ensemble of networks generated under some null model @xcite .
the typical null model prescribes an algorithm by which many randomized networks can be produced from the original network ( see , e.g. , milo et al .
@xcite for a review and comparison of several such algorithms ) .
while using an ensemble generated from the actual network often preserves features of the network that are desired for fair comparison ( e.g. the degree distribution ) , this method may also induce unintended correlations in subgraph counts that ultimately influence the labeling of subgraphs as motifs .
the purpose of this note is to demonstrate and interpret such correlations in a simple case and describe how mutual information may be used to identify such correlations in general . |
observations of accreting black holes of different mass , from the stellar mass black holes to the supermassive black holes in the center of galaxies , show impressive similarities of data , which point at identical physical processes in accreting plasma . also , black hole candidates demonstrate a great variety of physical conditions in the flows and , possibly , existence of a variety of accretion regimes .
existing theories describe different regimes of black hole accretion flows , which can be realized under different physical conditions . if matter accretes with a low specific angular momentum , @xmath1 , it forms spherical flows , which are described by bondi solution @xcite in the case of adiabatic flows . here
@xmath2 is the gravitational radius of the black hole , and @xmath3 is the speed of light .
if matter has a large specific angular momentum , @xmath4 , then accretion disks are formed .
the structure of accretion disks crucially depends on the efficiency of radiative cooling . if matter radiates efficiently , i.e. the radiative cooling time is shorter than the accretion time , @xmath5 , the disks are geometrically thin , @xmath6 , where @xmath7 is the scale height of the flow , and @xmath8 is the corresponding radius .
the radial structure of such disks is described by the shakura & sunyaev model @xcite . in the case of radiatively inefficient flows , @xmath9 ,
the internal energy of accreting matter is close to the virial energy , and the formed disks are geometrically thick , @xmath10 .
the thick disks can be formed in two limiting cases of very high , @xmath11 , and very low , @xmath12 , accretion rates , where @xmath13 is the eddington luminosity of the black hole . in the former case ,
the optical depth of accretion flow is very large , and photons , which carry most of the internal energy , are trapped inside the inflowing matter and can not be radiated away . in the latter case , the accreting plasma is very diluted , so that different radiative mechanisms are inefficient to cool plasma on the accretion time scales .
a model which describes the geometrically thick accretion disks was called the advection - dominated accretion flow ( adaf ) and attracted a considerable attention during the last two decades [ 310 ] .
the main feature of adafs is that most of the locally released gravitational and rotational energies of accretion flow is advected inward in the form of the gas internal energy , and the latter is finally absorbed by the black hole .
the recent interest to adafs was generated mostly by attempts using this model to explain the low - luminosity x - ray objects of both galactic and intergalactic nature ( for recent reviews see [ 11,12 ] ) .
properties of adafs could be better understood by analyzing the self - similar solution of the height - integrated hydrodynamical equations @xcite .
the solution depends on two parameters , the viscosity parameter @xmath14 and the adiabatic index @xmath15 , and satisfies the following scalings for the density @xmath16 , the radial velocity @xmath17 , the angular velocity @xmath18 , the isothermal sound speed @xmath19 and @xmath7 : @xmath20 @xmath21 @xmath22 @xmath23 @xmath24 where @xmath25 is the keplerian angular velocity . because of neglection of the radiative energy losses of the flow , the flow structure is independent of the mass accretion rate @xmath26 .
the accretion rate determines only scales of density through the relation @xmath27 .
two inconsistencies were found related to the model of adafs .
the first is that adafs must be convectively unstable [ 4,5,7 ] .
the second is connected to the positiveness of `` bernoulli parameter '' in adafs @xcite , @xmath28 where @xmath29 is the specific enthalpy .
a hypothesis was proposed that the latter problem could be solved by assuming the presence of powerful bipolar outflows in the accretion flows [ 13,14 ] . however , according to the later investigations [ 15,16 ] , the hypothesis met some difficulties .
the consequences of the earlier mentioned inconsistencies for adafs was understood after performing two- and three - dimensional hydrodynamic simulations of the inefficiently radiative accretion flows [ 1722 ] .
the simulations have revealed a new accretion regime of inefficiently radiated plasma , in which convection plays a dominant role determining structure and dynamics of the flow .
such a regime we shall refer as cdaf , the convection - dominated accretion flow .
behavior of convective blobs in non - rotating medium depends sensitively on the superadiabatic gradient @xmath30 which determines the brunt - visl frequency @xmath31 given by @xmath32 where @xmath33 is the radial effective gravity .
when @xmath34 is positive , perturbations of the blobs have an oscillatory behavior with the frequency @xmath31 and the medium is convectively stable .
however , when @xmath34 is negative , perturbations have a runaway growth , leading to convection .
convection is present whenever @xmath35 is positive , i.e. when the entropy gradient , given by @xmath36 is negative .
this is the well - known schwarzschild criterion . when there is rotation , a new frequency enters the problem and convection is no longer determined purely by the brunt - visl frequency . in this case , the effective frequency @xmath37 of convective blobs is given by @xmath38 where @xmath39 is the epicyclic frequency ; for @xmath40 , we have @xmath41 . again ,
when @xmath42 is negative , the rotating flows are convectively unstable .
when convection is developed , the convective blobs carry outward some amount of thermal energy . under certain approximations ( see details in @xcite ) the convective energy flux can be expressed in the form , @xmath43 where @xmath44 is the specific entropy , @xmath45 is the temperature , @xmath46 is the diffusion coefficient , defined by @xmath47 , and @xmath48 is the characteristic mixing - length .
the coefficient @xmath46 can also be expressed in the @xmath14-parameterization form , @xmath49 , where @xmath50 is a dimensionless parameter that describes the strength of convective diffusion ; this parameter is similar to the usual shakura & sunyaev @xmath14 . in the case of the self - similar adaf solution ( 1 ) , calculations lead to the conclusion that @xmath51 , i.e. adafs are convectively unstable and must have the outward energy fluxes defined by equation ( 2 ) .
to describe the results of numerical simulations of radiatively inefficient accretion flows , we shall mainly follow [ 21,22 ] . in these
works the accretion flows were simulated by solving the nonrelativistic , time - dependent navier - stokes equations with thermal conduction , @xmath52 @xmath53 @xmath54 where @xmath55 is the specific internal energy , @xmath56 is the gravitational potential of the black hole , @xmath57 is the viscous stress tensor with all components included , @xmath58 is the heat flux density due to thermal conduction and @xmath59 is the dissipation function . no radiative cooling was assumed and the ideal gas equation of state , @xmath60 , was adopted . only the shear viscosity was considered , @xmath61 where @xmath14 is a constant , @xmath62 . in the simulations
, it was assumed that mass is steadily injected into the calculation domain from an equatorial torus near the outer boundary .
matter is injected with almost keplerian angular momentum .
owing to viscous spread , a part of the injected matter moves inward and forms an accretion flow .
the computations were started from an initial state in which there is a small amount of mass in the domain . after a time comparable to the viscous timescale
, the accretion flow achieves a quasi - stationary behavior . to 0.0 in results obtained by using two - dimensional axisymmetric simulations
are summarized in figure 1 .
four types of accretion flows can be distinguished from the models with no thermal conduction ( see figure 1 , left panel ) .
convective flows . for a very small viscosity , @xmath63 ,
accretion flows are convectively unstable .
axially symmetric convection transports the angular momentum _
inward _ rather than outward .
convection governs the flow structure , making a flattened density profile , @xmath0 , with respect to the one for adafs .
convection transports a significant amount ( up to @xmath64 ) of the dissipated binding energy outward .
no powerful outflows are present .
\2 . large - scale circulations . for a larger , but still small viscosity , @xmath65 , accretion flows could be both stable or unstable convectively , depending on @xmath14 and @xmath15 .
the flow pattern consists of the large - scale ( @xmath66 ) meridional circulations .
no powerful unbound outflows are presented . in some respect
this type of flow is the limiting case of the convective flows in which the small - scale motions are suppressed by larger viscosity .
pure inflows . with an increasing viscosity , @xmath67 , the convective instability dies off .
some models ( with @xmath68 ) are characterized by a pure inflow pattern , and agree in many respects with the self - similar adafs .
no outflows are present .
bipolar outflows . for a large viscosity , @xmath69 ,
accretion flows differ considerably from the simple self - similar models .
powerful unbound bipolar outflows are present
. to 0.93 in figure 2 illustrates the density distribution and flow pattern in the low viscosity models .
the models show complicated time - dependent behavior with numerous vortices and circulations , and with density fluctuations . however , the time - averaged flow patterns are smooth and do not demonstrate small - scale features .
figure 3 shows typical behavior of radial profiles of variety of time- and angle - averaged variables in convective models . except near the inner and outer boundaries ( @xmath70 and @xmath71 )
, the profiles of @xmath16 , @xmath19 , @xmath72 and @xmath18 show prominent power - law behaviors .
two - dimensional models have demonstrated that the ( @xmath73,@xmath74)-component of the volume averaged reynolds stress tensor takes negative values in the case of convective flows .
this means that axisymmetric convection turbulence transports the angular momentum inward .
indeed , if there are no azimuthal gradients of pressure , the turbulence attempts to erase the angular momentum gradient , which for the keplerian - like angular velocity , @xmath40 , means that the angular momentum transports inward .
however , this is a property of axisymmetric flows .
does convection in real three - dimensional flows move angular momentum inward ?
do such flows have significant azimuthal perturbations which destroy axisymmetry ?
the answer to this question has been found with the help of fully three - dimensional simulations , without the limiting assumption of axisymmetry @xcite .
the simulations have demonstrated a good qualitative and quantitative coincidence of the results of two- and three - dimension modeling .
it was demonstrated that in the differentially rotating flows with nearly keplerian angular velocities , all azimuthal gradients are efficiently washed out , leading to almost axisymmetrical structure of three - dimensional flows .
figure 4 illustrates axisymmetric structure of the low - viscosity accretion flow obtained in three - dimensional simulations .
all low - viscosity models have negative volume - averaged @xmath75 in all ranges of the radii . only temporal convective blobs and narrow regions at the disk surfaces show positive @xmath75 .
it seems that the outward directed convection energy transport provides an additional cooling mechanism of the flows that always results in averaged @xmath76 , opposite to positive @xmath75 obtained in adafs . in several axisymmetric simulations , effects of turbulent thermal conduction
have been studied .
it was found that the conduction has an important influence on the flow structure , but it does not introduce a new type of flow in addition to those already discussed ( see figure 1 , right panel ) .
the conduction leads to a suppression of small - scale convection in the low - viscosity models . in the moderate - viscosity models the thermal conduction acts as a cooling agent in the outflows , reducing or even suppressing them .
obtained properties of cdaf can have important implications for the spectra and luminosities of accreting black holes [ 20,22,23 ] .
indeed , since @xmath77 and @xmath78 , the bremsstrahlung cooling rate per unit volume varies as @xmath79 .
after integration of @xmath80 over bulk of the flow one obtains the bremsstrahlung luminosity @xmath81 .
this means that cdafs mostly radiate on the outside , whereas most of the radiative energy of adafs comes from the innermost region @xcite .
numerical simulations of convective accretion flows have been understood qualitatively after construction of analytical self - similar solutions of such types of flows [ 24,25 ] . in the presence of convection
, the height - integrated angular momentum and energy equations can be written as follows , @xmath82 @xmath83 where @xmath84 , @xmath85 and @xmath86 are the advective , viscous and convective angular momentum fluxes respectively .
solutions of equations ( 5 ) and ( 6 ) , together with the equation of motion in the radial direction , crucially depend on the direction of convective angular momentum transport .
there are two extreme possibilities .
the first one is that convection behaves like normal viscosity , @xmath87 an alternative possibility is that the convective flux is directed down the specific angular momentum gradient , @xmath88.\ ] ] for @xmath40 , convective fluxes ( 7 ) and ( 8) correspond to outward and inward transport of angular momentum respectively .
numerical simulations have demonstrated that @xmath89 in cdafs .
choosing the convective flux ( 8) one can construct a _ nonaccreting _ solution with @xmath90 .
we refer to it as a `` convective envelope '' . for this solution , from equation ( 5 )
one obtains @xmath91 and @xmath92 .
the latter means that in the convective envelopes the net angular momentum flux vanishes : the inward convective flux is exactly balanced by the outward viscous flux .
because of absence of advection ( due to @xmath90 ) and zero local dissipation rate ( due to zero net angular momentum flux ) , the energy equation ( 6 ) is satisfied trivially with @xmath93 . finally , the convective envelope solution has the following scalings : @xmath94 @xmath95 @xmath22 @xmath23 @xmath96 the outward energy flux @xmath97 corresponds to luminosity @xmath98 , which is independent of radius .
source of this energy flux is formally located at @xmath99 , but the nature of the source is not specified in the self - similar solution . in more realistic cdafs ,
the mass accretion rate @xmath26 is small , but it is not exactly zero .
the fraction of the binding energy of the accreted mass released in the innermost region is the source of the energy required to support convection . in the case of cdafs
the convective luminosity can be expressed in the following form , @xmath100 , where the parameter @xmath101 has been estimated in numerical simulations , @xmath102 .
a non - zero @xmath26 leads to a finite @xmath103 .
the latter scaling and the scalings of other quantities in ( 9 ) agree quite well with the numerical results ( see figure 3 ) .
there is good reason to believe that `` viscosity '' in differentially rotating accretion flows is produced by magnetic stress generated by the magneto - rotational instability ( mri ) @xcite .
mhd simulations of accretion flows under conditions , in which cdafs could be formed , have been recently performed ( [ 2732 ] and matsumoto et al . in this proceedings ) .
a general conclusion of these studies is that mri , and probably other instabilities of magnetized medium , leads to development of turbulence in accretion flows .
the turbulence re - distributes angular momentum in the flows with an effective @xmath104 .
the simulations show a tendency for density profiles in accretion flows to be flatter , than those in adafs . in general , these results of mhd simulations do not contradict the results of hydrodynamical simulations of convective flows , which were discussed in previous sections .
the mhd simulations have demonstrated the range of @xmath105 , in which hydrodynamical models are convectively unstable .
the flattened density profiles in mhd models are in good agreement with those found in convective hydrodynamical models
. however , a direct comparison of results of both approaches meets difficulties .
the main problem is a low spatial resolution , and as a consequence , a small radial range of flow patterns studied in the discussed mhd simulations . in this case of
small radial ranges the effects of boundaries become significant enough to influence the flow structure .
thus , more powerful computers and more developed numerical codes are required to obtain a comparable spatial resolution to that used in hydrodynamical simulations .
another problem which can introduce difficulties for the comparison is that some mhd simulations have uncontrolled losses of energy due to numerical reconnection of magnetic lines .
these energy losses can artificially suppress convection in the considered flows and change the flow structures , independently of the used spatial resolution .
the problem could be solved by using an `` artificial '' resistivity in mhd schemes ( see @xcite for detailed discussion ) .
* radiatively inefficient hydrodynamical black hole accretion flows with small viscosity ( @xmath106 ) are always dominated by convection .
thus for low viscosity flows the concept of adaf is unphysical .
* at present , cdafs provide the best theoretical model for understanding of radiatively inefficient , low viscosity accretion flows . * cdafs are hot
, their thermal energy is close to the virial energy .
they have relatively reduced mass accretion rate . the density distribution in cdafs
is flattened , comparing to the one in adafs and bondi - like flows ( @xmath77 in the former case vs. @xmath107 in the latter case ) .
convection transports outward a significant amount ( up to @xmath64 ) of dissipated binding energy of accretion flows .
no powerful outflows are present in cdafs .
* cdaf is a very simple model which is now undergoing a rapid development .
a more detailed account of plasma , magnetic , and radiative processes is likely to change this model .
the authors thank nordita and i.v.i .
thanks harvard - smithsonian center for astrophysics for hospitality while part of this work was done .
this work was supported by nsf grant phy 9507695 , the royal swedish academy of sciences , and rfbr grant 00 - 02 - 16135 .
bondi h. , _ mnras _ , * 112 * , 195 ( 1952 ) .
shakura n.i . , and sunyaev r.a . , _ astron . &
_ , * 24 * , 337 ( 1973 ) . ichimaru s. , _ astrophys .
journal _ , * 214 * , 840 ( 1977 ) . gilham s. , _ mnras _ , * 195 * , 755 ( 1981 ) .
begelman m.c . , and meier d.l .
, _ astrophys .
journal _ , * 253 * , 873 ( 1982 ) .
abramowicz m.a . , czerny b. ,
lasota j .-
p . , and szuszkiewicz e. , _ astrophys .
journal _ , * 332 * , 646 ( 1988 ) .
narayan r. , and yi i. , _ astrophys . journal _ ,
* 428 * , l13 ( 1994 ) .
abramowicz m.a . , chen x. , kato s. , lasota j .-
p . , and regev o. , _ astrophys . journal _ , * 438 * , l37 ( 1995 ) .
narayan r. , and yi i. , _ astrophys . journal _
, * 444 * , 231 ( 1995 ) .
narayan r. , and yi i. , _ astrophys . journal _ , * 452 * , 710 ( 1995 ) .
kato s. , fukue j. , and mineshige s. , _ black - hole accretion disks _ , kyoto : kyoto univ . press ( 1998 ) .
abramowicz m.a . , blrnsson g. , and pringle j.e . ,
_ theory of black hole accretion disks _ , cambridge : cambridge univ . press ( 1998 ) .
xu g. , and chen x. , _ astrophys .
journal _ , * 489 * , l29 ( 1997 ) .
blandford r.d . , and begelman m.c . , _ mnras _ , * 303 * , l1 ( 1999 ) .
ogilvie g.i . ,
_ mnras _ , * 306 * , l9 ( 1999 ) .
abramowicz m.a . ,
lasota j .-
p . , and igumenshchev i.v . , _ mnras _ , * 314 * , 775 ( 2000 ) .
igumenshchev i.v . , chen x. , and abramowicz m.a . , _ mnras _ , * 278 * , 236 ( 1996 ) .
igumenshchev i.v . , and abramowicz m.a . ,
_ mnras _ , * 303 * , 309 ( 1999 ) .
stone j.m .
, pringle j.e . , and begelman m.c . , _ mnras _ , * 310 * , 1002 ( 1999 ) . igumenshchev i.v . , _ mnras _ , * 314 * , 54 ( 2000 ) .
igumenshchev i.v .
, abramowicz m.a . , and narayan r. , _ astrophys . journal _ , * 537 * , l27 ( 2000 ) .
igumenshchev i.v . , and abramowicz m.a . , _ astrophys .
journal sup .
_ , * 130 * , 463 ( 2000 ) .
ball g. , narayan r. , and quataert e. , _ astrophys .
, in press , astro - ph/0007037 .
narayan r. , igumenshchev i.v . , and abramowicz m.a .
, _ astrophys .
, * 539 * , 798 ( 2000 ) .
quataert e. , and gruzinov a. , _ astrophys .
journal _ , * 539 * , 809 ( 2000 ) .
balbus s.a . , and hawley j.f . , _ astrophys .
, * 376 * , 214 ( 1991 ) .
matsumoto r. , and shibata k. , in _ accretion phenomena and related outflows _
d. wickramsinghe , g. bicknell & l. ferrario , asp conf .
ser . 121 , 1997 , p. 443 .
matsumoto r. , in _ numerical astrophysics _
s. miyama , k. tomisaka & t. hanawa , dordrecht : kluwer , 1999 , p. 195 .
hawley j.f .
, _ astrophys .
journal _ , * 528 * , 462 ( 2000 ) .
machida m. , hayashi m.r . , and matsumoto r. , _ astrophys . journal _ ,
* 532 * , l67 ( 2000 ) .
machida m. , matsumoto r. , and mineshige s. , _ pasj _ , in press , astro - ph/0009004 .
stone j.m . , and pringle j.e . , _ mnras _ , in press , astro - ph/0009233 . | recent numerical simulations of radiatively inefficient accretion flows onto compact objects have shown that convection is a general feature in such flows .
dissipation of rotational and gravitational energies in the accretion flows results in inward increase of entropy and development of efficient convective motions .
convection - dominated accretion flows ( cdafs ) have a structure that is modified significantly in comparison with the canonical advection - dominated and bondi - like accretion flows .
the flows are characterized by the flattened radial density profiles , @xmath0 , and have reduced mass accretion rates .
convection transports outward a significant amount of the released binding energy of the accretion flow .
we discuss basic dynamical and observational properties of cdafs using numerical models and self - similar analytical solutions . |
protoplanetary disks evolve through a variety of physical processes . early in its evolution ,
a massive gas - rich disk loses mass through accretion onto the star , outflows , and photevaporation by high - energy photons ( gorti et al . 2009 ) . at the same time , grains grow into larger bodies that settle onto the mid - plane of the disk where they can grow into rocks , planetesimals , and beyond .
dust settling steepens the slope of the mid- and far - infrared ( ir ) spectral energy distribution ( sed ) because a smaller fraction of the stellar radiation is intercepted by circumstellar dust ( dullemond @xmath1 dominik , 2004 ) .
also , since grain growth is expected to proceed faster in the inner disk , it could result in an inner opacity hole ( dullemond @xmath1 dominik , 2005 ) .
as the disk mass and accretion rate decrease , chromospheric extreme - ultraviolet ( euv ) photons start to penetrate the inner disk and euv - induced photoevaporation becomes important .
when the accretion rate drops below the photoevaporation rate , the outer disk is no longer able to resupply the inner disk with material . at this point ,
the inner disk drains on a viscous timescale ( @xmath010@xmath2 yr ) and an inner hole is formed ( alexander et al . 2006 ) .
once this inner hole has formed , the euv photons can reach the inner edge of the disk unimpeded , preventing any material from the outer disk from flowing into the inner hole .
this halts accretion and results on the rapid transition between the classical t tauri star ( ctts ) and the weak - line t tauri star ( wtts ) stage .
the disk then quickly photoevaporates from the inside out .
once the remaining gas photoevaporates , the small grains are removed by radiation pressure and poynting - robertson drag .
what is left constitutes the initial conditions of a debris disk : a gas poor disk with large grains , planetesimals and/or planets .
this evolutionary path is certainly not unique as some systems such as lkca 15 , dm tau , and gm aur seem to have developed sharp cavities due to the dynamical interactions with close ( sub)stellar companions or recently formed giant planets while their outer disks are still quite massive ( najita et al . 2007 ) .
even though the broadest aspects of disk evolution summarized above are relatively well established ( see williams @xmath1 cieza , 2011 for a recent review on protoplanetary disks and their evolution ) , we are still far from developing a comprehensive disk evolution theory , for which additional observational constraints are much needed .
the so - called transition " objects ( broadly defined as disks with inner opacity holes ) are particularly useful _ disk evolution laboratories _ as they are the systems where the key physical processes mentioned above have the clearest observational signatures .
this is so simply because grain growth , photevaporation , and dynamical clearing all result in reduced levels of near- and/or mid - ir excess emission , which is the defining feature of transition objects .
this paper is the third part of a series from an ongoing project aiming to characterize @xmath0100 _ spitzer_-selected transition disks located in nearby star - forming regions .
this coordinated project has two main goals : 1 ) provide observational constraints on the evolution of primordial disks , their dissipation , and the primordial to debris disk transition , and 2 ) identify systems with strong evidence for ongoing giant planet formation so they can eventually be used as _
planet formation laboratories_. in the first paper of the series ( cieza et al . 2010 ; paper i hereafter ) , we studied a sample of 26 ophiuchus transition disks , while in the second paper ( romero et al .
2012 , apj , in press ; paper ii hereafter ) we analyzed a sample of 17 objects from the lupus , corona australis , and scorpius regions . in papers
i and ii , we showed that transition disks are a very heterogenous group of objects with disk masses ranging from @xmath3 0.6 m@xmath4 to 40 m@xmath4 , accretion rates ranging from @xmath3 10@xmath5 to 10@xmath6 m@xmath7 , and fractional disk luminosities , l@xmath8/l@xmath9 , ranging from a few percent to @xmath10 10@xmath11 .
we also showed that their diverse properties can be understood in terms of different disk evolution processes and distinct evolutionary stages . here we present millimeter wavelength photometry ( from the submillimeter array and jcmt ) , high - resolution optical spectroscopy ( from the canada - franca - hawaii telescope ) , and adaptive optics near - infrared imaging ( from the gemini telscope ) for other 31 _ spitzer_-selected transition circumstellar disks located in the perseus , taurus , and auriga molecular clouds .
as in our previous studies , we use these new ground - based data to estimate disk masses , accretion rates , and multiplicity for our sample in order to investigate the mechanisms potentially responsible for their inner opacity holes .
these new 31 objects take our total sample of well characterized transition disks to 74 .
the structure of this paper is as follows .
our sample selection criteria are presented in section 2 , while our observations and data reduction techniques are described in section 3 .
we present our results on disk masses , accretion rates , and multiplicity in section 4 . in section 5 ,
we discuss the likely origins of the inner holes of the 31 new individual targets and discuss the properties of our combined sample of 74 transition objects in a broader context of disk evolution .
finally , a summary of our results and conclusions is presented in section 6 .
we drew our sample from the _ spitzer _ catalogs of 3 northern clouds mapped by 3 different legacy programs : _ cores to disks _ ( perseus molecular cloud ) , _ taurus _ ( taurus molecular cloud ) , and _
gould s belt _ ( auriga molecular cloud ) . for a description of the _ cores to disks _ data products , see evans et [email protected] ] .
the _ taurus _
catalogs are discussed by padgett et al .. a the time of this writing , the _ gould s belt _ data products have not yet been delivered to nasa s infra - red science archive ( irsa ) , but they should also eventually become publicly available through irsa .
as in papers i and ii , we selected all the targets meeting the following criteria : _ a. _ have _ spitzer _ colors [ 3.6]-[4.5 ] @xmath3 0.25 .
these are objects with small or no near - ir excess ( see figure [ f : sample_sel ] ) .
the lack of a [ 3.6 ] - [ 4.5 ] color excess in our targets is inconsistent with an optically thick disk extending inward to the dust sublimation radius , and therefore implies the presence of an inner opacity hole .
the presence of this inner opacity hole is the defining feature we intend to capture in our sample . _
b. _ have _ spitzer _ colors [ 3.6]-[24 ] @xmath13 1.5 .
we apply this criterion to ensure that all our targets have very significant 24 @xmath14 m excesses ( @xmath135 - 10 @xmath15 ) , unambiguously indicating the presence of circumstellar material .
the exact color cut is somewhat arbitrary .
we have empirically found that most ysoc with low 24 @xmath14 m excesses ( [ 3.6]-[24 ] @xmath10 1.5 ) are in fact background agb stars ( see paper ii ) . _
c. _ are detected with a signal to noise ratio @xmath13 7 in all 2mass and irac wavelengths as well as at 24 @xmath14 m , to ensure that we only include targets with very reliable photometry . _
d_. have k@xmath16 @xmath3 11 mag , driven by the sensitivity of our near - ir adaptive optics observations and to ensure a negligible extragalactic contamination ( padgett et al .
2008 ) . _
e. _ are brighter than r = 18 mag according to the usno - b1 catalog ( monet et al .
2003 ) , driven by the sensitivity of our optical spectroscopy observations .
the first two selection criteria ( [ 3.6]-[4.5 ] @xmath3 0.25 and [ 3.6]-[24 ] @xmath13 1.5 ) effectively become our working definition for a transition disk .
these criteria are fairly inclusive and encompass most of the transition disk definitions found in the literature as they select disks with a significant flux decrement relative to normal disks " in the near - ir or at all wavelengths . in particular
, our definition includes objects with falling mid - ir seds such as anemic ( lada et al .
2006 ) or homologously depleted transition disks ( currie et la .
2009 ) as well as objects with rising mid - ir seds such as classical transition disks ( muzerolle et al .
2010 ) or cold disks ( brown et al . 2007 ; merin et al .
the one type of transition disk that is likely to be under - represented in our sample is the so - called pre - transitional " disk category , which describes systems with optically thin gaps separating optically thick inner and outer disk components ( espaillat et al .
these rare objects tend to have large near - ir excesses that could be excluded by our [ 3.6]-[4.5 ] @xmath3 0.25 criterion . for a concise description of the complex transition disk nomenclature , see evans et al .
we find 41 targets that meet all of our selection criteria : 18 objects in perseus ( @xmath044@xmath17 ) , 6 in auriga ( @xmath015@xmath17 ) , and 17 in taurus ( @xmath041@xmath17 ) .
we have observed all 41 objects in our target list ; however , as discussed in section [ pms_id ] , this list includes 6 asymptotic giant branch ( agb ) stars and 4 likely debris disks around main - sequence ( ms ) stars .
the remaining 31 targets are bona fide young stellar objects ( ysos ) with circumstellar disks and constitute our science sample .
the 2mass ids and alternative names , 2mass and _ spitzer _ fluxes , and the usno - b1 r - band magnitudes for all our 41 targets are listed in table 1 .
the spitzer data of our perseus transition disks have been presented by jorgensen et al .
( 2006 ) and rebull et al .
similarly , the parent sample of our taurus objects has been discussed by rebull et al .
( 2010 ) and luhman et al .
we note that while rebull et al . only used the catalogs from the taurus legacy program as we do , the study by luhman et al .
includes many additional observations of the taurus region from several general observer and guaranteed time observations programs .
some well - known taurus transition disks included in luhman et .
( e.g. , dm tau and coku tau/4 ) were not observed as part of the taurus legacy program and therefore are not included in our sample .
the _ spitzer _
data for the parent population of our auriga targets have not been published at the time of this writing .
we obtained echelle spectroscopy for our entire sample using the espadons echelle spectrograph on the 3.5-meter canada - france - hawaii telescope ( cfht ) at mauna kea observatory in hawaii .
the observations were performed in service mode over 4 semesters ( 2009a , 2009b , 2010a , and 2010b ) .
the spectra were obtained in the standard
star+sky " mode , which delivers the complete optical spectra between 3500 and 10500 at a resolution of 68,000 , or 4.4 km / s . for each object
, we obtained a set of 3 spectra with exposures times ranging from 2.5 to 30 minutes each , depending on the brightness of the target .
the data were reduced through the standard cfht pipeline upena , which is based on the reduction package libre - esprit . in section [ results ] , we use these data to derive the spectral types and accretion rates of our targets .
high spatial resolution near - ir observations of our entire sample were obtained in service mode during the 2009b semester with the 8.1-meter gemini north telescope in mauna kea using the near infrared imager and spectrometer ( niri ) and the altair adaptive optics ( ao ) system .
we used the natural guide star mode for the targets brighter than r = 15.5 mag , and the laser guide star mode for the objects fainter than r=15.5 mag . in all cases ,
the science target was used as the tip / tilt guide star .
j- and k - band observations were obtained with the f/32 camera ( 0.022 arcsec / pix , 22@xmath18 @xmath19 22@xmath18 field of view ) .
the brightest objects required narrow - band filters . for each target , we took five dithered images in each filter with individual exposure times ranging from 1 to 20 sec .
the k - band ao images are the most stable and provide the highest strehl ( @xmath040@xmath17 ) , and thus allow the detection of binaries with large brightness ratios .
the j - band images provide a psf core with a smaller fwhm ( @xmath00.05@xmath18 ) and are better suitable for the detection of very tight systems .
the data were reduced with image reduction and analysis facility ( iraf ) using the nireduce task in the niri package . in section [ multi_sec ]
, we use these ao imaging data to constraint the multiplicity of our targets .
the ( sub)millimeter wavelength photometry is the most expensive component of our observing program in terms of observing time , preventing us from obtaining such data for the entire sample .
two of our targets , # 2 and 26 ( fw tau ) , have already been detected at ( sub)millimeter wavelengths ( merin et al .
2010 ; andrews & williams , 2005 ) , while stringent upper limits exist for # 33 ( zz tau ; andrew & williams , 2005 ) .
we also found 850 @xmath14 m submillimeter array ( sma ; ho et al .
2004 ) archival data for object # 9 .
we have observed 26 of the remaining 37 objects with the sma .
we also obtained common - user bolometer array-2 ( scuba-2 ) data using the james clerk maxwell telescope ( jcmt ) for 5 of the 11 targets that were not observed with the sma .
six objects remain unobserved at ( sub)millimeter wavelengths . however , since our ( sub)millimeter observing programs gave the lowest priority to objects with little 24 @xmath14 m excesses and/or lacking accretion signatures , these 6 unobserved targets turned out to be either an asymptotic giant branch ( agb ) stars contaminating our sample ( 1 object ) or a debris disk candidate ( 5 objects ; see section [ pms_id ] ) .
we note that debris disks are not expected to be detectable at the sensitivity levels of our ( sub)millimeter wavelength survey , m@xmath20 @xmath0 6 - 15 m@xmath21 ( corresponding to @xmath02 - 5 m@xmath4 for primordial disks with a gas to mass dust ratio of 100 ) .
in fact , this dust mass level is an order of magnitude higher than that of the most massive debris disks observed ( wyatt 2008 ) .
none of our conclusions are thus likely to be affected by the lack of a complete millimeter wavelength data set .
our sma observations were conducted in service mode , during the 2008b , 2009b , and 2010b semesters . virtually all the observations were obtained in the compact configuration and with the 230 ghz/1300 @xmath14 m receiver .
the only exception was object # 31 , which was observed in the extended configuration using the 345 ghz/850 @xmath14 m receiver .
both the upper and lower sideband data were used , resulting in a total bandwidth of 4ghz .
typical zenith opacities during our 230 ghz observations were @xmath22 @xmath00.150.25 .
the 345 ghz extended configuration observations were obtained under significantly better conditions ( @xmath22 @xmath00.07 ) . for each target , the observations cycled between the target and two gain calibrators ( either 3c111 and 3c84 or 0336 + 323 and 0449 + 113 ) , with 20 - 30 minutes on target and 7.5 minutes on each calibrator .
the raw visibility data were calibrated with the mir reduction packagecqi / mircook.html ] .
the passband was flattened using @xmath01 hour scans of 3c454.3 or 3c279 .
the absolute flux scale was determined through observations of either uranus or titan and is estimated to be accurate to 15@xmath17 .
the flux densities of detected sources were measured by fitting a point source model to the visibility data using the uvmodelfit task in the common astronomy software applications package casaobtaining.shtml ] , while upper limits were derived from the rms of the visibility amplitudes .
the rms noise of our sma observations range from @xmath00.4 to 2.0 mjy per beam .
we detected , at the 3@xmath15 level or better , 4 of the sma targets : # 5 , 29 , and 31 ( from our own data ) and object # 9 from the archival data .
objects # 2 and 26 from the literature take the number of ( sub)millimeter detections to 6 .
the ( sub)millimeter wavelength fluxes ( and 3-sigma upper limits ) for our sample are listed in table 2 . in section [ disk_mass_sec ]
, we use the ( sub)millimeter wavelength photometry to constrain the masses of our transition disks .
the scuba-2 data were obtained in service mode on february 28 and march 1 , 2010 during shared - risk observations using only one of the 4 arrays per channel of the final scuba-2 instrument ( holland et al .
each of the 5 objects observed with scuba-2 ( targets # 6 , 8 , 37 , 38 , and 40 ) were scanned following a daisy pattern for 15 to 25 minutes . during both nights
the weather was quite good ( @xmath23 = 0.06 - 0.08 ) ; however , both the 450 and 850 @xmath14 m arrays were significantly noisier than expected during shared - risk observations .
the strong quasars 3c84 and 3c111 were used as callibrators .
the data was reduced using the submm user reduction facility ( smurf ) .
none of the scuba-2 targets were detected in the final mosaics , where the rms noise at 850 @xmath14 m is in the @xmath05 to 10 mj range the 450 @xmath14 m mosaics are much noisier ( rms @xmath24 100 mjy ) and have little value for our program .
the 3-@xmath15 upper limits at 850 @xmath14 m for our scuba-2 targets are listed in table 2 .
we estimate the spectral types of our targets by comparing temperature sensitive features in our echelle spectra to those in templates from stellar libraries .
we use the libraries presented by soubiran et al .
( 1998 ) and montes ( 1998 ) .
the former has a spectral resolution of 42,000 and covers the entire 4500 6800 spectral range .
the latter has a resolution of 12,000 and covers the 4000 9000 spectral range with some gaps in the coverage . before performing the comparison ,
we normalize all the spectra and take the template and target to a common resolution .
m - type stars were assign spectral types based on the strength of the tio bands centered around 6300 , 6700 , 7 and 7150 .
we classify g - k stars based on the ratio of the v i ( at 6199 ) to fe i ( 6200 ) line ( padgett , 1996 ) and/or on the strength of the ca i ( 6112 a ) and na i ( 5890 and 5896 ) absorption lines ( montes et al . 1999 ; wilking et al .
the f , a , and b stars were identified and typed by the width of the underlying h@xmath25 absorption line ( which is much wider than its emission line ) and/or by the strength of the paschen 16 , 15 , 14 , and 13 lines .
the spectral types so derived are listed in table 2 .
we estimate the typical uncertainties in our spectral classification to be 1 spectral subclass for m - type stars and 2 spectral subclasses for k and earlier type stars .
spectral types for most of our targets in perseus ( the ic 348 members ) and taurus are presented in luhman et al .
( 2003 ) , rebull et al .
( 2010 ) , and/or luhman et al .
we find that most of our spectral types do in fact agree within 1 or 2 spectral subclasses with previously published values .
background objects are known to contaminate samples of _ spitzer_-selected yso candidates ( harvey et al .
2007 , oliveira et al . 2009
; papers i and ii ) . at low flux levels ,
background galaxies are the main source of contamination .
however , the optical and near - ir flux cuts we have implemented as part of our sample selection criteria are very efficient at removing extragalactic sources from the _ spitzer _ catalogs . at the bright end of the flux distribution , asymptotic giant branch ( agb ) stars are the main source of contamination .
agb stars are surrounded by shells of dust and thus have small , but detectable , ir excesses .
the extreme luminosites ( @xmath010@xmath26 l@xmath27 ) of agb stars imply that they can pass our optical and near - ir flux cuts even if they are located several kpc away .
agb stars can easily be identified in our sample as late m - type stars lacking both h@xmath25 emission ( from either accretion or chromospheric activity ) and li i 6707 absorption .
we find 6 such objects in our sample , all in the taurus region .
their full optical spectra are shown in figure [ f : agb_spt ] .
as seen in figure [ f : sample_sel ] , the contamination of agb stars is particularly high in the [ 3.6]-[24 ] colors range between 1.5 and 2.5 ( five of the six agb stars fall in this narrow range ) .
the li i 6707 absorption line is a very good indicator of stellar youth in mid - k to m - type stars because li is burned very efficiently in the convective interiors of low - mass stars and is depleted soon after these objects arrive on the main - sequence ( cargile et al .
2010 ) ; however , it is not an age discriminant for early type stars where the depletion timescales can approach the main - sequence ages . to establish whether our early type targets are in fact consistent with pre - main - sequence ( pms ) stars associated with the perseus , taurus , and auriga molecular clouds , we place them in the hertzsprung
russell ( h - r ) diagram ( see figure [ f : hrd ] ) and compare their position against the theoretical isochrones from siess et al .
we adopted the bolometric corrections and temperature scale from kenyon @xmath1 hartmann ( 1995 ) and the following distances : 320 pc for perseus ( herbig 1998 ) , 140 pc for taurus ( kenyon , gmez & whitney , 2008 ) , and 300 pc for auriga ( heiderman et al .
we corrected for extinction using a@xmath28 = 1.53 @xmath19 e(j - k@xmath29 ) , where e(j - k@xmath30 ) is the observed color excess with respect to the expected photospheric color for the given spectral type ( also from kenyon @xmath1 hartmann 1995 ) .
we find that 5 of our early type ( f5 to b9 ) targets fall below the 10 myr isochrone ( # 11 , 30 , 35 , 36 , and 38 ) .
all of them have very low disk luminosities and could be background ms stars with debris disks .
ms stars in this temperature range are 2.5 to 100 times more luminous than the sun and can be seen at large distances .
their large luminosities also imply that their debris disks are more likely to be detectable at _ spitzer _ wavelengths because the fractional disk luminosity needed to produce a detectable 24 @xmath14 m excess is much lower for baf - type stars than it is for lower luminosity objects ( cieza et al .
however , object # 11 in perseus has been identified as a member of ic 348 cluster ( luhman et al .
2003 ) based on proper motion measurements . given the uncertainties associated with placing objects in the h - r diagram , we consider object # 11 to be a pms star . in section [ classification ] , we classify the other four under - luminous objects as ms debris disk candidates .
the h - r diagram also serves as an independent check for our sample of late - type targets . as seen in figure [
f : hrd ] , all the k and m - type objects fall within the 0.5 and 10 myr isochrones , except for 5 of the 6 agb stars which are clearly too over - luminous to be low - mass objects at the distance of the taurus molecular cloud . setting aside 6 agb stars ( targets # 25 , 27 , 34 , 37 , 39 and 40 ) , and 4 likely ms stars ( sources # 30 , 35 , 36 , and 38 ) we are left with 31 objects that constitute our sample of transition disks around bona fide pms stars . unlike papers i and ii , where we found that virtually all the objects were k and m - type stars , here we find that @xmath030@xmath17 of the bona fide transition objects ( 9/31 ) are bafg - type stars . this difference is likely to be due to a combination of several effects .
first , the initial mass function ( imf ) is not the same in all the clouds .
lupus and ophiuchus contain a larger proportion of very low - mass stars ( i.e. , late m - type objects ) than taurus ( hughes et al .
1994 ; erickson et al .
second , many low - mass objects at the distance of perseus and auriga ( 300 - 320 pc ) are too faint to meet all of our sample selection criteria .
finally , there are significant differences in the sizes of the pms populations .
since there are @xmath03 more yso candidates in perseus than in lupus , perseus should contain a larger number of high mass stars for a given imf .
most young low - mass stars show h@xmath25 emission , either from chromospheric activity or accretion .
non - accreting objects show narrow ( @xmath10 200 km / s , measured at 10 @xmath17 of the peak intensity ) and symmetric line profiles of chromspheric origin , while accreting objects present broad ( @xmath24 270 km / s ) and asymmetric profiles produced by large - velocity magnetospheric accretion columns . in order to measure the velocity width of the h@xmath25 emission line , we first subtract the continuum by fitting a line to the spectrum in the velocity intervals @xmath31600 to @xmath31400 km / s and 400 to 600 km / s centered at the h@xmath25 location .
the peak of the continuum - subtracted spectrum is then normalized to 1 , and the velocity width , @xmath32 , is measured at 10 @xmath17 of the peak intensity .
the velocity dispersion so derived for the h@xmath25 emission lines of our transition disks are listed in table 2 .
we estimate the typical uncertainty in @xmath32 to be of the order of 10@xmath17 .
however , as discussed below , the uncertainty in the corresponding accretion rate is expected to be much larger .
the boundary between accreting and non - accreting objects has been empirically placed by different studies at @xmath32 between 200 km / s ( jayawardana et al . 2003 ) and 270 km / s ( white @xmath1 basri , 2003 ) . since only two objects , sources # 20 and # 23
have @xmath32 in the 200 - 270 km / s range , most accreting and non - accreting objects are clearly separated in our sample .
source # 20 is a m6 star with a rising mid - ir sed , a 70 @xmath14 m detection , and @xmath32 @xmath0 210 km / s .
very low - mass stars tend to have narrower h@xmath25 lines than higher mass objects because of their lower accretion rates ( nata et al .
2004 ) and their lower gravitational potentials ( muzerolle et al .
2003 ) . given all the available data , we classify target # 20 as an accreting object , but warn the reader that its accreting nature is less certain than that of the rest of the objects classified as cttss .
similarly , source # 23 has @xmath32 @xmath0200 km / s and we consider it to be non - accreting because of its very low fractional disk luminosity ( l@xmath33/l@xmath9 @xmath3 10@xmath34 , see [ sed_mor ] ) , and it is most likely an optically thin debris disks ; however , its non - accreting nature is not certain .
the continuum - subtracted h@xmath25 profiles for all the 13 accreting transition disks are shown in figure [ f : prof_acc ] .
there are 9 k and m - type stars in our sample with narrow h@xmath25 emission consistent with chromospheric activity .
the continuum - subtracted h@xmath25 profiles of these objects are shown in figure [ f : prof_non_acc ] .
the spectra of targets # 13 and 16 have lower signal to noise ratios than the rest of the spectra in the figure , which makes it difficult to measure accurate @xmath32 values .
target # 13 is an m2 star with a small h@xmath25 equivalent widths ( 3.5 ; luhman et al .
2003 ) and object # 16 has l@xmath33/l@xmath9 @xmath3 10@xmath34 and it is another likely debris disk . for disk classification purposes ,
we consider both objects to be non - accreting .
the 13 bafg stars in our sample show the characteristic photospheric h@xmath25 line in absorption , with little or no evidence for significant superimposed h@xmath25 emission . as discussed in the previous section , 4 of them are likely to be background ms star ( objects # 30 , 35 , 36 , and 38 ) .
the other 9 objects are consistent with non - accreting pms stars and are bona fide targets . for accreting objects
, the velocity dispersion of the h@xmath25 line correlates well with accretion rates derived from models of the magnetospheric accretion process .
we therefore estimate the accretion rates of our targets from the width of the h@xmath25 line measured at 10@xmath17 of its peak intensity , adopting the relation given by natta et al .
( 2004 ) : @xmath35 this relation is valid for 600 km / s @xmath13 @xmath32 @xmath13 200 km / s ( corresponding to 10@xmath6 m@xmath27/yr @xmath13 m@xmath36 10@xmath5 m@xmath27/yr ) and can be applied to objects with a range of stellar ( and sub - stellar ) masses .
as discussed by muzerolle et al .
( 2003 ) and sicilia - aguilar et al .
( 2006b ) , the broadening of the h@xmath37 line is 1 to 2 orders of magnitude more sensitive to _ low _ accretion rates than other accretion indicators such as u - band excess and continuum veiling measurements , and is thus particularly useful to distinguish weakly accreting from non - accreting objects .
however , the 10@xmath17 width measurements are also dependent on the line profile , rendering the 10@xmath17 h@xmath25 velocity width a relatively poor _ quantitative _ accretion indicator ( nguyen et al .
2009 ) .
for the objects we consider to be non - accreting , we adopt a mass accretion upper limit of 10@xmath5m@xmath27/yr , corresponding to @xmath32 = 200 km / s , although the detectability of accretion is both a function of spectral type and data quality .
the so derived accretion rates ( and upper limits ) for our sample of transition disks are listed in table 3 .
given the large uncertainties associated with equation 2 and the intrinsic variability of accretion in pms stars , these accretion rates should be considered to be order - of - magnitude estimates .
disk masses obtained from modeling the ir and ( sub)millimeter seds of circumstellar disks are well described by a simple linear relation between ( sub)millimeter flux and disk mass ( andrews @xmath1 williams 2005 ; 2007 ) .
following papers i and ii , we adopt the linear relations presented by cieza et al .
( 2008b ) : @xmath38 m_{jup } \label{eq_mass}\ ] ] @xmath39 m_{jup } \label{eq_mass}\ ] ] these relations come from the ratios of model - derived disk masses to observed ( sub)millimeter fluxes presented by andrews @xmath1 williams ( 2005 ) for 33 taurus stars .
based on the standard deviation in the ratios of the model - derived masses to observed ( sub)millimeter fluxes , the above relation gives disk masses that are within a factor of @xmath02 of model - derived values ; nevertheless , much larger _
systematic _ errors can not be ruled out ( andrews @xmath1 williams , 2007 ) . in particular , the models from andrews @xmath1 williams , 2005 ; 2007 ) assume an opacity as a function of frequency of the form @xmath40
@xmath41 @xmath42 and a normalization of @xmath43 = 0.1 gr/@xmath44 at 1000 ghz .
this opacity implicitly assumes a gas to dust mass ratio of 100 .
both the opacity function and the gas to dust mass ratio are highly uncertain and expected to change due to disk evolution processes such as grain growth and photoevaporation .
detailed modeling and additional observational constraints on the grain size distributions ( e.g. , from sub / millimeter wavelength slopes ) and the gas content ( e.g. , from co , [ o i ] , and/or [ c ii ] observations ) will be needed to derive more accurate disk masses for each individual transition disk . the disk masses ( and 3-@xmath15 upper limits ) for our sample are listed in table 3 and were derived adopting the distances to the clouds from section [ pms_id ] .
the vast majority of our transition objects have estimated disk masses lower than @xmath013 m@xmath4 .
however , 5 of them have disk masses typical of cttss ( @xmath0315 m@xmath4 ) . from the visual inspection of our gemini - ao images
, we identify 13 multiple systems : targets # 1 , 4 , 5 , 14 , 15 , 18 , 22 , 24 , 26 , 28 , 31 , 32 , and 33 ( see figure [ f : multi ] ) .
the separation and flux ratios of these systems range from 0.05@xmath18 to 1.7@xmath18 and 1.0 to 14 , respectively .
object # 1 is a triple .
most systems were fully resolved in both our j- and k@xmath29-band images , which have typical fwhm values of 0.06@xmath18 - 0.08@xmath18 .
the tightest systems , targets # 26 and 33 , were only fully resolved by the j - band images ( fwhm @xmath0 0.05@xmath18 ) . in perseus ,
targets # 14 and 15 are previously known binaries ( duchene et al .
1999 ) , while targets # 1 , 4 , 5 , and 18 are newly identified multiple systems .
similarly , objects # 24 , 26 , and 33 are known taurus binaries ( koponacky et al . 2007 ; simon et al .
1995 ) , while targets # 22 is a newly discovered taurus binary system .
all 3 multiples in auriga ( objects # 28 , 31 , and 32 ) are newly identified systems . for target # 32
, the ao system did not lock correctly in either of the two components of the binary , resulting in a very poor ao correction and a much uncertain determination of the flux ratio .
for the apparently single stars , we estimated the detection limits at 0.1 and 0.2@xmath18 separations from the 5-@xmath15 noise of psf - subtracted images .
since no psf standards were observed in our program , we subtract a psf constructed by azimuthally smoothing the image of the target itself , as follows . for each pixel in the image , the separation from the target s centroid
is calculated , with sub - pixel accuracy .
the median intensity at that separation , but within an arc of 30 pixels in length , is then subtracted from the target pixel .
thus , any large scale , radially symmetric structures are removed .
the separations , positions angles , and flux ratios of the multiple systems in our sample are shown in table 2 .
the flux ratio limits for unseen companions at 0.1 and 0.2@xmath18 separations , obtained as described above , are also listed for the targets that appear to be single .
in addition to the disk mass , accretion rate , and multiplicity , the sed morphology and fractional disk luminosity of a transition disk can provide important clues on the nature of the object .
following papers i and ii , we quantify the sed morphologies and fractional disk luminosities of our transition objects and use these quantities as part of our disk classification scheme ( see section [ classification ] ) .
we quantify the sed shape " of our targets adopting the two - parameters introduced by cieza et al .
( 2007 ) : @xmath45 , which is the longest wavelength at which the observed flux is dominated by the stellar photosphere , and @xmath46 , the slope of the ir excess , computed as dlog(@xmath47f)/dlog(@xmath47 ) between @xmath45 and 24 @xmath14 m . to calculate @xmath45
, we compare the extinction - corrected sed to nextgen models ( hauschildt et al .
1999 ) normalized to the j - band and choose @xmath45 as the longest wavelength at which the stellar photosphere contributes over 50@xmath17 of the total flux .
the uncertainty of @xmath45 is roughly one sed point .
the @xmath45 and @xmath46 values for our entire sample are listed in table 3 .
the @xmath45 and @xmath46 parameters are dependent on the sed sampling and are affected by ir variability ( espaillat et al . 2011 ) ; however , they provide first - order information on the structure of the disk .
for a given stellar luminosity , the @xmath45 value correlates with the size of the inner hole as it depends on the temperature of the dust closest to the star .
it is clear however , that a given @xmath45 value implies a much larger inner hole for a disk around an a - type star than for one around a m - type star .
similarly , @xmath46 correlates well with the sharpness of the opacity hole . on the one hand
, sharp inner holes result in positive @xmath46 values , which are typical of classical transition disks ( muzerolle et al .
2010 ) and cold disks ( brown et al . 2007 ; merin et al .
2010 ) with rising mid - ir seds . on the other hand ,
more radially continuous disks that have undergone significant grain growth and dust settling show the negative @xmath46 values ( dullemond @xmath1 dominik , 2004 ) that is characteristic of anemic ( lada et al .
2006 ) and homologously depleted ( currie et la .
2009 ) transition disks with falling mid - ir seds .
see espaillat et al .
( 2012 ) for a recent discussion on the link between sed morphology and physical properties of the disk .
the fractional disk luminosity , the ratio of the disk luminosity to the stellar luminosity ( l@xmath8/l@xmath9 ) , is another important quantity that relates to the evolutionary state of a disk .
typical primordial disks around cttss have l@xmath8/l@xmath9 @xmath0 0.1 as they have optically thick disks that intercept ( and reemit in the ir ) @xmath010@xmath17 of the stellar radiation .
in contrast , debris disks show l@xmath8/l@xmath9 values @xmath10 10@xmath34 because they have optically thin disks that intercept and reprocess @xmath010@xmath4810@xmath34 of the star s light ( bryden et al . 2006 ) .
we estimate l@xmath8/l@xmath9 for our sample of disks by integrating over frequency the flux contribution of the disk and the star to the observed sed ( see paper i for calculation details ) .
the near and mid - ir luminosities of our disks are well constrained because their seds are relatively well sampled at these wavelengths .
however , their far - ir luminosities remain more poorly constrained .
only 5 of our 41 targets have ( 5-@xmath15 or better ) detections at 70 @xmath14 m listed in the _ c2d _ , _ goulds belt _ , and _ taurus _ catalogs .
for the rest of the objects , we have obtained 5-@xmath15 upper limits as in paper i ( from the noise of the 70 @xmath14 m images at the location of the targets ) in order to fill the gap in their seds between 24 @xmath14 m and the millimeter .
the log(l@xmath33/l@xmath49 ) values for our transition disk sample , ranging from @xmath311.7 to @xmath314.9 are listed in table 3 .
the log(l@xmath33/l@xmath49 ) values are highly dependent on sed sampling and should be considered order of magnitude estimates . as most of the luminosity of a disk extending inward to the dust sublimation temperature is emitted in the near - ir , l@xmath33/l@xmath9 is a very strong function of @xmath45 : the shorter the @xmath45 wavelength , the higher the fractional disk luminosity . for objects with @xmath45 @xmath3 8.0 @xmath14 m , the 70 @xmath14 m flux represent only a minor contribution to the total disk luminosity . on the other hand ,
objects with with @xmath45 @xmath50 8.0 @xmath14 m ( i.e. , objects where the ir excess only becomes significant at 24 @xmath14 m ) have much lower l@xmath33/l@xmath49 values , and the 70 @xmath14 m emission becomes a much larger fraction of the total disk luminosity . as a result
, the l@xmath33/l@xmath49 values of objects with @xmath45 @xmath50 8.0 @xmath14 m and no 70 @xmath14 m detections should be considered upper limits .
one of the main results from papers i and ii has been the very large range of disk properties ( accreation rates , disks masses , sed morphologies , and l@xmath33/l@xmath49 values ) exhibited by our targets .
this diversity points toward different evolutionary stages and/or different physical processes driving the evolution of each disk .
in particular , the wealth of information discussed in the previous sections allow us to place each target in our transition sample into the following categories : grain - growth dominated disks , giant planet - forming disks , photoevaporating disks , debris disks , and circumbinary disks .
accreting objects with @xmath46 @xmath3 0 ( falling mid - ir seds ) are most readily explained as primordial disks that have undergone significant grain growth and dust settling .
the seds of the 7 transition disks in this category are shown in figure [ f : sed_ggd ] .
circumstellar disks are initially very flared and intersect a significant fraction of the stellar radiation , most of which is reprocessed and reemitted at ir wavelengths . as dust grains coagulate and
grow in the disk , they fall toward the mid - plane , where the surface density is higher and they can grow at a higher rate and settle even deeper into the disk . as a result , the disk becomes geometrically flatter with time , which reduces the fraction of stellar radiation intercepted by the disk and steepens the slope of the sed at mid - ir wavelengths ( dalessio et al . 2006 ; dullemond et al .
grain growth is expected to proceed faster in the inner disk , where the surface densities are higher and the dynamical timescales are shorter .
thus , the depletion of micron - sized grains in the inner disk through grain growth also contributes to the low levels of near and mid - ir excesses seen in this type of objects ( dullemond @xmath1 domink , 2005 ) .
based on their seds , grain - growth dominated disks could also be classified as anemic ( lada et al .
2006 ) or homologously depleted transition disks ( currie et la .
they show a large diversity of accretion rates and disk masses , which in this paper range from 10@xmath51 to 10@xmath52 m@xmath27 yr@xmath53 and @xmath3 0.5 to 5.3 m@xmath4 ( but note doar 25 in paper i , with a disk mass of @xmath040 m@xmath4 and an accretion rate of 10@xmath54 m@xmath27 yr@xmath53 ) . establishing whether the decrement of ir excess is mostly due to a reduction of the dust opacity , a reduction of the surface density in the disk , or purely geometrical effects requires resolved observations and/or details modeling of individual objects . in some cases
, the weakness of the ir excess might be attributed to the very low luminosity of the central star ( see section 5.2.4 ) . " accreting transition disks with sharp inner hole ( i.e. , @xmath46 @xmath24 0 ; rising mid - ir seds ) are currently considered the most likely sites for ongoing planet formation ( najita et al . 2007 ; merin et al .
2010 ; williams & cieza , 2011 ) . the recent identification of planet candidates within the inner holes of the t cha ( hulamo et al .
2011 ) and lkca 15 disks ( kraus @xmath1 ireland , 2011 ) strongly supports this interpretation .
we find that 6 objects fall in this category ( see figure [ f : sed_pfd ] ) .
the presence of accretion unambiguously identify them has gas - rich primordial disks .
the connection between a rising mid - ir seds and the presence of a sharp inner hole ( a large change in surface density over an small radial distance ) is well established through both sed modeling ( brown et al .
2007 ; calvet et al . 2002 ; 2005 ) and direct submillimeter imaging ( brown et al . 2009 ; hughes et al . 2007 ; 2009 ) . while
in addition to planet formation , other processes such as grain growth , photoeveporation , and dynamical clearing due to _ stellar _ companions have been proposed to explain this type of objects , these alternative explanations face serious theoretical and/or observational challenges . in particular ,
grain growth efficiency is expected to be an smooth function of radius , which is inconsistent with the abrupt change in opacity inferred for transition disk with steeply rising mid - ir seds . whether photoevaporation can account for the inner holes of accreting objects depends on the photoevaporation rates . as will be discussed in section [ photo_constrain ] ,
there is strong evidence _ against _ photoevaporation rates being large enough to explain the inner holes of accreting transition disks .
we thus conclude that the presence of accretion makes the photoevaporation scenario much less likely . dynamical clearing by ( sub)stellar companions ( lubow & dangelo , 2006 ; zhu et al . 2011 )
therefore remains the most likely explanation for accreting transition disks with rising mid - ir seds .
recent hydrodynamical simulations of multiple planets embedded within the disk also help reconciling the lack of near - ir excess with the levels of accretion seen in transition disks in this category .
the optically thick but physically narrow tidal tails predicted by these simulations can transport significant amounts of material without over - predicting the ir excesses observed ( dodson - robinson & salyk , 2011 ) . while our ao observations can not rule out the presence of stellar companions to our targets inward of ( @xmath015 - 30 au ) , near - ir interferometry ( pott et al . 2010 ) and aperture masking observations ( kraus et al .
2009 ) conclusively exclude _ stellar _ companions as a possible cause of the inner holes in several similar objects . all things considered ,
we conclude that the dynamical clearing by one or multiple planets is the most likely explanation for the properties of the 6 objects shown in figure [ f : sed_pfd ] .
we find that virtually all non - accreting objects remain undetected at millimeter wavelengths , implying dust masses below 610 m@xmath21 ( corresponding to disk masses @xmath10 23 m@xmath4 for a gas - rich disk with a gas to dust mass ratio of 100 ) .
target # 26 ( fw tau ) with an estimated disk mass of 0.4 m@xmath4 is the only exception .
all photoevaporation models predict the formation of a gap in the disk and the subsequent draining of the inner disk once most of the outer disk mass has been depleted and the accretion rate falls below the photoevaporation rate .
once this inner hole has formed , the high - energy photons can reach the inner edge of the disk unimpeded , and the disk quickly dissipates from the inside out . after the circumstellar gas photoevaporates ,
the small grains are removed by radiation pressure and poynting - robertson drag , leaving behind a gas - poor disk with large grains , planetesimals and/or planets ( i.e. , a debris disk ) . in this context ,
the most fundamental difference between a photoevaporating disk and a young debris disk is the presence of primordial gas at large radii : if the inner hole is due to photoevaporation , the outer disk should remain gas - rich beyond the photoevaporation front .
since we lack information on the gas content in the outer disks of our targets , we adopt a less direct criterion based on the fractional disk luminosity , l@xmath33/l@xmath49 , to tentatively distinguish primordial photoevaporating disks from debris disks .
as discussed in section [ sed_mor ] , typical primordial disks around cttss have l@xmath8/l@xmath9 @xmath0 0.1 as they have optically thick disks that reprocess @xmath010@xmath17 of the stellar radiation , while debris disks show l@xmath8/l@xmath9 values @xmath10 10@xmath34 because their optically thin disks only intercept @xmath010@xmath4810@xmath34 of the star s light ( bryden et al .
disks around young wttss have l@xmath8/l@xmath9 values that fill the gap between those two regimes suggesting they are an evolutionary link between the two stages ( padgett et al .
2006 ; cieza et al . 2007 ; wahhaj et al . 2010 ) .
following papers i and ii , we classify as photoevaporating disks objects with l@xmath8/l@xmath9 @xmath24 10@xmath34 and as debris disks objects with l@xmath8/l@xmath9 @xmath10 10@xmath34 .
the seds of our 7 photevaporating disk candidates are shown in figure [ f : sed_photo ] , while the seds of our 15 debris disk candidates are shown in figure [ f : sed_debris ] .
the central objects of 4 of the debris disk candidates appear to be significantly underluminous in the h@xmath31r diagram ( targets # 30 , 35 , 36 , and 38 ) and could be background ms stars ( see section [ pms_id ] ) .
co observations with the atacama large millimeter / submillimeter array ( alma ) will be able to measure the gas content of all non - accreting disks and conclusively establish their primordial or debris disk status .
early multiplicity surveys of pms stars have shown that most stars in the solar neighborhood form in multiple systems ( leinert et al .
1993 ; ghez et al . 1993 ; simon et al . 1995 ) .
since most of these binary systems have orbits with semimajor axes of the order of the typical sizes of circumstellar disks , the dynamical interaction of the stellar components with each other s disks has dramatic effects on disk devolution .
the outer disks around the individual stars in a binary system are expected to be tidally truncated at a fraction ( @xmath00.5 ) of the binary separation .
similarly , the circumbinary disk , if present , should have an inner radius @xmath02@xmath19 the semi - major axis of the system ( artymowicz & lubow , 1994 ) . while it is not the only possible outcome , tidal truncation in binary systems
is known to produce inner opacity holes resulting in transition disk seds .
such is the case of the famous coku tau/4 system ( ireland @xmath1 kraus , 2008 ) . in section [ multi_sec ]
, we found that 13 of our targets are in fact multiple systems .
we now discuss the likelihood of each system to retain a circumbinary disk responsible for the observed seds .
nine of the binary systems in our sample ( objects # 4 , 5 , 14 , 15 , 18 , 22 , 28 , 31 , and 32 ) have _ projected _ separations greater than 100 au ( see table 3 ) .
these projected separations represent the minimum value of the current physical separation .
any circumbinary disk around such wide systems should have an inner hole @xmath24 200 au and would most likely remain undetectable in our seds .
targets # 1 , 24 , and 33 have companions at 62 , 90 , and 10 au , respectively .
however , they are all m - type stars with significant excesses at 8 @xmath14 m micron , implying the presence of dust at separations of the order of a few au or less . also , their seds show no evidence for dynamically induced inner holes .
we thus conclude that their transition disk status is not a result of their stellar companions .
object # 26 ( fw tau ) , on the other hand , has a projected separation of only 7 au and a weak 24 @xmath14 m excess and is therefore fully consistent with a circumbinary disk .
the fw tau system shares some of the properties of coku tau/4 , such as its very low disk mass ( @xmath10 0.5 m@xmath4 ) and the lack of detectable accretion .
the lower luminosity of fw tau ( it is a m5 star ) might explain its lower mid - ir excess with respect to coku tau/4 even if their inner holes are of similar size .
we classify fw tau as both a circumbinary and a photoevaporating disk candidate as tidal truncation and photoevaporation are not mutually exclusive processes .
an important caveat of the circumbinary disk classification is the lack of constraints for most of the sample on stellar companions with separations @xmath1010 au , a range where @xmath030@xmath17 of all stellar companions are expected to be found ( duquennoy @xmath1 mayor , 1991 ; kraus et al .
future radial velocity and aperture masking observations are likely to increase the number of objects in the circumbinary disk category . combining our sample of 31 transition disks with those from papers i and ii results in a sample of 74 homogenously selected and characterized transition disk objects . in this section ,
we discuss the properties of the combined sample and the important clues transition disks can provide on different aspects of disk evolution .
we find that only @xmath018@xmath17 ( 13/74 ) of the transition disks in our combined sample have properties that are best explained by the dynamical interaction of recently formed giant planets .
calculating the overall incidence of planet - forming disk candidates among protoplanetary disks in nearby molecular clouds is non - trivial because the fraction of transition disks and agb contamination vary greatly from cloud to cloud . we calculate this important quantity as follows : we exclude from our calculation the taurus molecular cloud because its catalog does not provide a ysoc classification , which results in planet - forming disk fraction of 21.0@xmath17 ( 12/57 ) among the remaining well - characterized transition disks .
from table 3 in paper ii , we find that these 57 objects have been drawn from a sample of 1059 ysoc in 8 different molecular clouds ( lup i , iii , iv , v and vi , ophiuhcus , perseus , and auriga ) , of which 24.8@xmath17 ( 263/1059 ) satisfy our main sample selection criteria ( [ 3.6]-[4.5 ] @xmath3 0.25 and [ 3.6]-[24 ] @xmath13 1.5 ) .
we thus estimate the overall fraction of planet - forming disk candidates among ysoc to be [email protected] = 5.2@xmath17 .
the ysoc catalogs are known to be contaminated by agb stars , and this contamination is higher among objects with transition disk seds ( agb typically have photosphere near - ir colors and small 24 @xmath14 m excesses ) . from the estimates in paper ii s table 3 , we find that the overall ysoc contamination by agbs in the 8 clouds we consider is 10.3@xmath17 , while the contamination increases to 15.6@xmath17 in the region of the color - color diagram occupied by transition disks .
correcting the statistics from agb contamination decreases the incidence of transition disks in the ysoc sample to 23.3@xmath17 ( 222/950 ) . similarly , the overall fraction of planet - forming disk candidates among ysoc decreases slightly to [email protected] = 4.9@xmath17 . for comparison ,
the incidence of giant planets within 20 au of mature solar - type stars is estimated to be @xmath020@xmath17 ( cumming et al .
this discrepancy strongly suggests that not all giant planets embedded in a primordial disk result in clear observational signatures .
the presence of hidden planets " is supported by the results from recent hydrodynamic simulations showing that multiple giant planets are needed in order to have a detectable effect in the emerging sed ( zhu et al . 2011 ; dodson - robinson & salyk 2011 ) .
furthermore , high - resolution ( sub)millimeter images have revealed inner holes in systems that have perfectly `` normal '' seds ( andrews et al .
this can easily be understood considering that ( sub)millimeter observations can detect modest reductions in the surface density of the disk , while reducing the levels of the near and mid - ir excess emission requires an extreme depletion of small dust particles . also
, the planet formation signature capture by our selection criteria is expected to be present for only a small fraction of the disk lifetime , after the planets have become massive enough to dynamically open a gap in the disk and the inner disk has drained . in other words , many protoplanetary disks with
normal " seds may contain one or more protoplanets that are not yet massive enough to open a wide hole in the disk , and as mentioned in section 2 , we are likely to miss pre - transition disks , with wide gaps but significant near - ir excess . we thus conclude that the giant planet disk candidates identified in our survey are likely to represent only a subsample of the entire population of disks actively forming planets in the molecular clouds we are considering . this subsample may be dominated by systems with multiple and/or very massive planets .
dynamical clearing by stellar companions has been one of the main mechanisms proposed to explain the inner holes of transition disks .
however , in section [ classification ] we only found one circumbinary disk candidate in our sample of 31 transition disks .
a low incidence of circumbinary disk candidates was also found in papers i and ii .
the combined sample of 74 transition disks now allow us to derive more robust conclusions on the role multiplicity plays on transition disk systems . as in paper
i , we compare the distribution of binary separations of our sample of transition disk to that of non - transition disk and disk - less pms stars .
we draw our sample of non - transition disks ( objects with normal " levels of near - ir excess ) and disk - less pms stars from the compilation of multiplicity and _ spitzer _ data ( for disk identification ) presented by cieza et al .
( 2009 ) for over 300 pms stars in the taurus , chameleon i , ophiuchus , and corona australis regions .
as seen in figure [ f : multi_fig ] , disk - less pms stars tend to have companions at smaller separations than stars with regular , non - transition disks . according to a two - sided kolmogorov
smirnov ( ks ) test , there is less than a 10@xmath11 probability that the distributions of binary separations of non - transition disks and disk - less stars have been drawn from the same parent population .
this result can be understood in terms of the effect tidal truncation has on the lifetimes of the circumstellar disks of the individual components of the binary system ( cieza et al .
2009 ; kraus et al .
2011 ) . a binary system at the peak of the separation distribution ( @xmath030 au ) , is expected to initially have individual disks that are @xmath010 - 15 au in radius .
given that the viscous timescale is roughly proportional to the size of the disk , such small disks are likely to have accretion lifetimes smaller than the age of the sample and hence they now appear as young disk - less stars .
in very tight binary systems , the outer disk can survive in the form of a circumbinary disk , with a tidally truncated inner hole with a radius @xmath02@xmath19 the orbital separation ( artymowicz @xmath1 lubow , 1994 ) .
if such close systems were a significant component of the transition disk population , one would expect to find a higher incidence of close binaries in transition disks that in non - transition disks . while our multiplicity census is clearly very incomplete at small separations
, our ao observations should be sensitive to an over abundance of companions in the 10 to 30 au range that could in principle be responsible for many of the inner holes of our transition disks .
however , we find that only @xmath04@xmath17 ( 3/74 ) of our transition disks have companions in this separation range .
near - ir interferometry ( pott et al . 2010 ) and aperture masking observations ( kraus et al .
2009 ) also suggest that the incidence of tight _ stellar _ companions is rather low in transition objects .
we thus conclude that tight stellar companions tend to destroy each other s disk rather quickly .
few close binary systems retain circumbinary disks with transition disk seds ( e.g. , coku tau/4 and fw tau ) .
while it is increasingly clear that photoevaporation by the central star plays a fundamental role on the evolution and dissipation of protoplanetary disks , the heating mechanisms and magnitudes of the photoevaporation rates are still a matter of intense debate .
in particular , the relative importance of far - uv ( fuv ) , extreme - uv ( euv ) , and x - ray photoevaporation is not well understood .
all photoevaporation models predict the formation of an inner hole and subsequent inside - out dissipation once mass transport across the disk falls below the photoevaporation rate ; however , they strongly disagree on the mass a disk should have at the time the inner hole is formed .
euv - driven photoevaporation models ( alexander et al .
2006 ) predict low evaporation rates , of the order of 10@xmath55 m@xmath27/yr . as a result , the inner hole forms late on the evolution the disk , when most of the mass has been depleted ( disk mass @xmath10 1 m@xmath4 ) and the accretion rate onto the star has dropped below the 10@xmath55 m@xmath27/yr level .
more recent euv+x - ray and fuv+x - ray photoevaporation models ( gorti et al . 2009 ; owen et al . 2011 ) predict much higher evaporation rates ( 10@xmath5610@xmath57 m@xmath27/yr ) and an earlier formation for the inner hole , when disk masses are @xmath24 5 m@xmath4 .
these later models allow the presence of accretion in disks with inner holes that have formed through photoevaporation because the accretion rates from the draining of the inner disk become detectable for significant period of time .
however , once the inner disk is drained , accretion must stop as material from the _ outer _ disk can not overcome the photoevaporation front .
x - ray photevaporation models thus predict a population of non - accreing objects ( i.e. , wttss ) with relatively massive outer disks ( @xmath24 5 m@xmath4 ) . in paper ii , we demonstrated the lack of massive outer disk around low - mass wttss , which contradicts the predictions of photoevaporating models with high accretion rates .
this result stills holds for the new low - mass stars studied herein and seems to extend to higher mass objects .
figure [ f : mass_vs_spt ] shows the disk masses derived for our combined sample of transition disks .
there is only one ( sub)millimeter detection at 0.4 m@xmath4 for non - accreting objects ( corresponding to fw tau ) , the rest are upper limits in the @xmath00.5 to @xmath05 m@xmath4 range . since fw tau is a circumbinary disk , there is not guarantee that its inner hole is due to photoevaporation .
the sample now includes several non - accreting kgfab stars observed at millimeter wavelengths , but increasing the sample size for this type of objects is clearly desirable .
an important caveat to figure [ f : mass_vs_spt ] is the inherent uncertainty of deriving disk masses from ( sub)millimeter photometry , which requires strong assumptions on dust opacities and gas to dust mass ratios ( williams @xmath1 cieza , 2011 ) .
if those assumptions are incorrect , they could lead to systematic errors in the determination of disk masses .
however , the observed accretion rates of pms stars provide independent observational constraints that also favor photoevaporations rates significantly lower than 10@xmath57 m@xmath27/yr .
if photoevaporation rates are @xmath010@xmath57 m@xmath27/yr , then it becomes difficult to explain all the cttss without inner holes that have accretion rates less than @xmath010@xmath57 m@xmath27/yr ( hartmann et al .
overall , the observational evidence suggests that photoevaporation rates must be small ( @xmath010@xmath55 m@xmath27yr@xmath53 ) and that the phoptoevaporation front can only overcome accretion when disk masses have fallen below @xmath10 1 m@xmath4 .
while the samples in papers i and ii were strongly dominated by m - type stars , our current combined sample does contain a significant number of higher mass kgfab - type stars .
these allow us to investigate whether the incidence of the disk clearing mechanisms depend on stellar mass or luminosity .
figure [ f : hrd_types ] shows the same h - r diagram as in figure [ f : hrd ] , but now indicating the location of each type of disk .
the most striking feature of the diagram is that all 9 pms stars hotter than @xmath010@xmath58 ( 5754 k , corresponding to a g5 star ) have non - accreting disks , either photoevaporating disks or debris disks .
this is in agreement with the results from _
spitzer _ studies of young stellar clusters and associations showing that primordial disks dissipate faster around higher mass stars ( carpenter et al .
2006 ; dahm @xmath1 hillenbrand , 2007 ) .
the 4 ms stars with debris disks are also hot baf - type stars .
this can be understood considering that debris disks are easier to detect at mid - ir wavelengths around more luminous objects ( e.g. , rieke et al .
2005 ; hernandez et al .
2006 ; cieza et al .
table 4 shows the occurrence of different types of transition disks for m - type ( 48 objects ) , k - type ( 17 objects ) and gfab - type stars ( 9 objects ) in our combined sample . since m - type stars are fainter and cooler than higher mass stars , they may present weaker near- and/or mid - ir excesses even without grain growth and dust settling effects ( ercolano , clarke @xmath1 robitaille , 2009 ) .
however , the fact that we see a _ lower _ fraction of grain - growth dominated disk candidates around m - type stars than around k - type stars , suggests that we are not significantly overestimating the fraction of m - type stars in this category .
detailed modeling of individual objects are still needed to confirm the nature of some of these grain - growth dominated disks , especially the ones around very late m - type stars with modest decrements of ir emission .
however , it is clear that grain - growth and dust settling play a fundamental role on the evolution of cicumstellar disks
. the grain - growth dominated disk category accounts for @xmath2440@xmath17 of all disks around both the m - type stars and k - type stars in our sample .
currie @xmath1 sicilia - aguilar ( 2011 ) recently showed that the percentage of disks in the transition phase increases significantly between @xmath01 to @xmath08 myr . the h - r diagram in figure [ f : hrd_types ] , showing our combined sample , now allows us to investigate how the incidence of different transition disk _ types _ evolve with age .
while the stellar ages of individual targets are highly uncertain , the age distribution of each disk category should carry more meaningful information .
table 5 shows the occurrence of different types of transition disks for objects falling above the 1 myr isochrone ( 14 targets ) , between the 1 and 3 myr isochrones ( 30 objects ) , and below the 3 myr isochrone ( 30 objects , _ not _ including the 4 ms star candidates ) in our combined sample . even with our relatively small sample , we find that the incidence of photoevaporating disks and debris disks increases with age , while the fraction of grain - growth dominated disks decreases with time .
these trends are in agreement with the overall evolution of typical protoplanetary disks ( williams @xmath1 cieza , 2011 ) .
interestingly , the occurrence of plant - forming disks candidates peaks in the 1 - 3 myr old age bin and is @xmath00@xmath17 for objects above the 1 myr age isochrone . the lack of planet - forming disks with ages @xmath10 1 myr is suggestive , especially considering that the intrinsic age distribution of these objects is most likely to be narrower than that seen in the h - r diagram , where some of the scatter can be attributed to the observational uncertainties in t@xmath59 and luminosity .
if taken at face at value , these results would imply that _ 1 ) _ the inferred giant planets have formed through core - accretion ( lissauer , 1993 ) as gravitational instability models ( boss , 2000 ) favor a younger age distribution , and _ 2 ) _ core accretion takes 2 to 3 myr to form giant planets massive enough to open a gap in the disk .
while important caveats remain ( we are dealing with planet - forming disk _ candidates _ , the sample is relatively small , and stellar ages are highly uncertain and model dependent ) , these results are promising . in the near future ,
detailed age analyses of larger samples of transition disks with increasingly clear planet - formation signatures are likely to provide strong astrophysical constraints on planet - formation timescales .
as part of an ongoing program aiming to characterize a large number of _ spitzer_-selected transition disks , we have obtained millimeter wavelength photometry , high - resolution optical spectroscopy and adaptive optics near - infrared imaging for a sample of 31 transition objects located in the perseus , taurus , and auriga molecular clouds .
we use these ground - based data to estimate disk masses , multiplicity , and accretion rates in order to investigate the mechanisms potentially responsible for their inner holes .
we combined disk masses , accretion rates , and multiplicity data with other information , such as sed morphology and fractional disk luminosity to classify the disks as _
strong candidates _ for the following categories : grain - growth dominates disks ( 7 objects ) , giant planet - forming disks ( 6 objects ) , photoevaporating disks ( 7 objects ) , debris disks ( 11 obecjts ) , and cicumbinary disks ( 1 object , which was also classified as photoevaporating disk ) .
each category represents an educated guess , giving all the available data , on the evolutionary status of the disk or the physical process mainly responsible for the reduced levels of near - ir and/or mid - ir excesses characteristic of the objects in our sample .
the boundaries between the categories are of course not perfectly defined .
for instance , all primordial disks are expected to simultaneously undergo some degree of grain growth , dust settling , and photoevaporation .
similarly , our criterion to distinguish primordial photoevaporating disks from debris disks is based on l@xmath8/l@xmath9 instead of on the gas content of their outer disks , which still remains highly unconstrained .
the gas dissipation of the outer disk through photoevaporation is believed to mark the rapid transition from the primordial to the debris disk stage ( williams @xmath1 cieza , 2011 ) .
conclusively establishing the nature of each target will thus require detail modeling and followup observations .
even with the above caveats , the properties of transition disks can provide important clues on different aspects of disk evolution . combining our sample of 31 transition disks with those from papers i and ii results in a sample of 74 transition disk objects that have been selected , characterized , and classified in an homogenous way .
the main conclusions derived from the analysis of this combined high - quality sample can be summarized as follows : \1 ) circumstellar disks with reduced levels of near - ir and/or mid - ir excesses in nearby molecular clouds represent a very heterogenous group of objects with a wide range of sed morphologies , disk masses , accretion rates , and fractional disk luminosities .
this diversity points toward distinct evolutionary stages and physical processes driving the evolution of each disk .
\2 ) the incidence of objects with signatures of dynamical clearing by recently formed giant planets is significantly lower than the occurrence of giant planets within 20 au of mature solar - type stars ( @xmath05@xmath17 vs @xmath020@xmath17 ) .
the giant planet disk candidates identified in our survey are likely to represent special cases , where multiple massive planets may be present .
\3 ) the incidence of circumbinary disk candidates in our sample of transition objects is low ( @xmath10 10@xmath17 ) implying that tight _ stellar _ companions tend to erode each other s disk rather quickly .
\4 ) there is a lack of massive disks around non - accreting stars of a wide range of spectral types , which contradicts the predictions of recent photoevaporation models that find very high evaporation rates ( @xmath010@xmath57 m@xmath27yr@xmath53 ) .
our results suggest that photoevaporation rates must be small ( @xmath010@xmath55 m@xmath27yr@xmath53 ) and that the photoevaporation front can only overcome accretion when disk masses have fallen below @xmath10 1 m@xmath4 .
5 ) debris disks and photoevaporating disk candidates are more common around hotter stars , consistent with the idea that primordial disks dissipate faster around more massive objects .
\6 ) grain growth - dominated disks account for @xmath2440@xmath17 of our sample of transition disks around k and m - type stars , confirming that grain - growth and dust settling play a major role on the evolution of primordial circumstellar disks .
\7 ) we find a trend in the sense that the incidence of photoevaporating disks and debris disks increases with age , while the fraction of grain - growth dominated disks decreases with time , which is consistent with disk evolution models .
\8 ) a preliminary analysis of the age distribution of disks with signatures of dynamical clearing by recently formed giant planets reveals a lack of such objects among the youngest stars in the sample .
this favors core accretion as the main planet formation mechanism and a 2 to 3 myr formation timescale .
transition objects are invaluable disk evolution and planet formation laboratories .
detailed modeling and followup observations of different types of transition disks are highly desirable to further our understanding of key processes such as grain growth and dust settling , photoevaporation , dynamical clearing , and planet formation itself .
we thank the anonymous referee whose comments and suggestions helped us to significantly improve the paper .
support for this work was provided by nasa through the _ sagan _ fellowship program under an award from caltech .
m.r.s thanks for support from fondecyt ( 1061199 ) and basal cata pfb 06/09 .
g.a.r . was supported by alma fund grant 31070021 .
m.d.m . was supported by alma - conicyt fund grant 31060010 .
j.p.w . acknowledges support from the national science fundation grant ast08 - 08144 .
arm acknowledges financial support from fondecyt in the form of grant number 3110049 .
this work makes use of data obtained with the _ spitzer _ space telescope , which is operated by jpl / caltech , under a contract with nasa .
alexander , r. d. , clarke , c. j. , & pringle , j. e. 2006 , , 369 , 229 andrews , s. m. , & williams , j. p. 2005
, , 631 , 1134 andrews , s. m. , & williams , j. p. 2007
, , 671 , 1800 andrews , s. m. , wilner , d. j. , espaillat , c. , et al .
2011 , , 732 , 42 artymowicz , p. , & lubow , s. h. 1994 , , 421 , 651 boss , a. p. 2000 , , 536 , l101 bryden , g. , et al .
2006 , , 636 , 1098 brown , j. m. , et al .
2007 , , 664 , l107 brown , j. m. , blake , g. a. , qi , c. , et al .
2009 , , 704 , 496 calvet , n. , dalessio , p. , hartmann , l. , wilner , d. , walsh , a. , & sitko , m. 2002 , , 568 , 1008 calvet , n. , et al .
2005 , , 630 , l185 cargile , p. a. , james , d. j. , & jeffries , r. d. 2010 , , 725 , l111 carpenter , j. m. , mamajek , e. e. , hillenbrand , l. a. , & meyer , m. r. 2006 , , 651 , l49 cieza , l. , et al .
2007 , , 667 , 308 cieza , l. a. , cochran , w. d. , & augereau , j .- c .
2008a , , 679 , 720 cieza , l. a. , swift , j. j. , mathews , g. s. , & williams , j. p. 2008b , , 686 , l115 cieza , l. a. , et al .
2009 , , 696 , l84 cieza , l. a. , schreiber , m. r. , romero , g. a. , et al .
2010 , , 712 , 925 cumming , a. , butler , r. p. , marcy , g. w. , et al .
2008 , , 120 , 531 currie , t. , lada , c. j. , plavchan , p. , et al .
2009 , , 698 , 1 currie , t. , & sicilia - aguilar , a. 2011 , , 732 , 24 dahm , s. e. , & hillenbrand , l. a. 2007 , , 133 , 2072 dalessio , p. , calvet , n. , hartmann , l. , franco - hernndez , r. , & servn , h. 2006 , , 638 , 314 dodson - robinson , s. e. , & salyk , c. 2011 , , 738 , 131 duchne , g. , bouvier , j. , & simon , t. 1999 , , 343 , 831 dullemond , c. p. , & dominik , c. 2004 , , 421 , 1075 dullemond , c. p. , & dominik , c. 2005 , , 434 , 971 dullemond , c. p. , hollenbach , d. , kamp , i. , & dalessio , p. 2007
, protostars and planets v , 555 duquennoy , a. , & mayor , m. 1991 , , 248 , 485 ercolano , b. , clarke , c. j. , & robitaille , t. p. 2009
, , 394 , l141 erickson , k. l. , wilking , b. a. , meyer , m. r. , robinson , j. g. , & stephenson , l. n. 2011 , , 142 , 140 espaillat , c. , calvet , n. , dalessio , p. , hernndez , j. , qi , c. , hartmann , l. , furlan , e. , & watson , d. m. 2007 , , 670 , l135 espaillat , c. , furlan , e. , dalessio , p. , et al .
2011 , , 728 , 49 espaillat , c. , ingleby , l. , hernandez , j. , et al .
2012 , arxiv:1201.1518 evans , n. , et al .
2009 , arxiv:0901.169 ] fiorucci , m. , & munari , u. 2003 , , 401 , 781 ghez , a. m. , neugebauer , g. , & matthews , k. 1993 , , 106 , 2005 gorti , u. , dullemond , c. p. , & hollenbach , d. 2009 , , 705 , 1237 hartmann , l. , calvet , n. , gullbring , e. , & dalessio , p. 1998
, , 495 , 385 harvey , p. , mern , b. , huard , t. l. , rebull , l. m. , chapman , n. , evans , n. j. , ii , & myers , p. c. 2007 , , 663 , 1149 hauschildt , p. h. , allard , f. , & baron , e. 1999 , , 512 , 377 heiderman , a. , evans , n. j. , ii , allen , l. e. , huard , t. , & heyer , m. 2010 , , 723 , 1019 herbig , g. h. 1998 , , 497 , 736 hernndez , j. , briceo , c. , calvet , n. , et al . 2006 , , 652 , 472 hulamo , n. , lacour , s. , tuthill , p. , et al .
2011 , , 528 , l7 ho , p. t. p. , moran , j. m. , & lo , k. y. 2004 , , 616 , l1 holland , w. , macintosh , m. , fairley , a. , et al .
2006 , , 6275 , hughes , j. , hartigan , p. , krautter , j. , & kelemen , j. 1994 , , 108 , 1071 hughes , a. m. , wilner , d. j. , calvet , n. , et al .
2007 , , 664 , 536 hughes , a. m. , andrews , s. m. , espaillat , c. , et al .
2009 , , 698 , 131 ireland , m. j. , & kraus , a. l. 2008 , , 678 , l59 jayawardhana , r. , mohanty , s. , & basri , g. 2003 , , 592 , 282 jensen , e. l. n. , cohen , d. h. , & gagn , m. 2009 , , 703 , 252 jrgensen , j. k. , harvey , p. m. , evans , n. j. , ii , et al .
2006 , , 645 , 1246 kenyon , s. j. , & hartmann , l. 1995 , , 101 , 117 kenyon , s. j. , gmez , m. , & whitney , b. a. 2008 , handbook of star forming regions , volume i , 405 kraus , a. l. , ireland , m. j. , martinache , f. , & lloyd , j. p. 2009
, american institute of physics conference series , 1094 , 453 kraus , a. l. , & ireland , m. j. 2011 , arxiv:1110.3808 kraus , a. l. , ireland , m. j. , hillenbrand , l. a. , & martinache , f. 2011 , arxiv:1109.4141 konopacky , q. m. , ghez , a. m. , rice , e. l. , & duchne , g. 2007 , , 663 , 394 lada , c. j. 1987 , star forming regions , 115 , 1 lada , c. j. , et al .
2006 , , 131 , 1574 leinert , c. , zinnecker , h. , weitzel , n. , et al .
1993 , , 278 , 129 lissauer , j. j. 1993 , , 31 , 129 lubow , s. h. , seibert , m. , & artymowicz , p. 1999
, , 526 , 1001 lubow , s. h. , & dangelo , g. 2006 , , 641 , 526 luhman , k. l. , stauffer , j. r. , muench , a. a. , et al .
2003 , , 593 , 1093 luhman , k. l. , allen , p. r. , espaillat , c. , hartmann , l. , & calvet , n. 2010 , , 186 , 111 mern , b. , brown , j. m. , oliveira , i. , et al .
2010 , , 718 , 1200 monet , d. g. , et al .
2003 , , 125 , 984 montes , d. , ramsey , l. w. , & welty , a. d. 1999 , , 123 , 283 muzerolle , j. , hillenbrand , l. , calvet , n. , briceo , c. , & hartmann , l. 2003 , , 592 , 266 muzerolle , j. , allen , l. e. , megeath , s. t. , hernndez , j. , & gutermuth , r. a. 2010 , , 708 , 1107 najita , j. r. , strom , s. e. , & muzerolle , j. 2007 , , 378 , 369 natta , a. , testi , l. , muzerolle , j. , randich , s. , comern , f. , & persi , p. 2004
, , 424 , 603 nguyen , d. c. , scholz , a. , van kerkwijk , m. h. , jayawardhana , r. , & brandeker , a. 2009 , , 694 , l153 oliveira , i. , et al .
2009 , , 691 , 672 owen , j. e. , ercolano , b. , clarke , c. j. , & alexander , r. d. 2010 , , 401 , 1415 padgett , d. l. 1996 , , 471 , 847 padgett , d. l. , cieza , l. , stapelfeldt , k. r. , et al .
2006 , , 645 , 1283 padgett , d. l. , et al .
2008 , , 672 , 1013 pott , j .- u . , perrin , m. d. , furlan , e. , et al . 2010 , , 710 , 265 rebull , l. m. , stapelfeldt , k. r. , evans , n. j. , ii , et al .
2007 , , 171 , 447 rebull , l. m. , padgett , d. l. , mccabe , c .- e . ,
2010 , , 186 , 259 rieke , g. h. , su , k. y. l. , stansberry , j. a. , et al .
2005 , , 620 , 1010 rromero , g. et al .
2012 , apj , in press sicilia - aguilar , a. , et al . 2006a , , 638 , 897 sicilia - aguilar , a. , hartmann , l. w. , frsz , g. , henning , t. , dullemond , c. , & brandner , w. 2006b , , 132 , 2135 siess , l. , dufour , e. , & forestini , m. 2000 , , 358 , 593 simon , m. , et al .
1995 , , 443 , 625 soubiran , c. , katz , d. , & cayrel , r. 1998 , , 133 , 221 wahhaj , z. , cieza , l. , koerner , d. w. , et al .
2010 , , 724 , 835 white , r. j. , & basri , g. 2003 , , 582 , 1109 wilking , b. a. , meyer , m. r. , robinson , j. g. , & greene , t. p. 2005
, , 130 , 1733 williams , j. p. , & cieza , l. a. 2011 , , 49 , 67 wyatt , m. c. 2008 , , 46 , 339 zhu , z. , nelson , r. p. , hartmann , l. , espaillat , c. , & calvet , n. 2011 , , 729 , 47 rrcrrrrrrrrrrr [ sample ] 1 & 03292681 + 3126475 & mbo 15 & 12.37 & 7.41e+01 & 1.03e+02 & 8.96e+01 & 4.72e+01 & 3.20e+01 & 2.33e+01 & 1.70e+01 & 1.89e+01 & @xmath3 8.06e+01 & per + 2 & 03292925 + 3118347 & mbo 22 & 16.56 & 1.47e+01 & 2.89e+01 & 2.75e+01 & 1.36e+01 & 1.02e+01 & 7.62e+00 & 6.94e+00 & 9.52e+01 & 2.91e+02 & per + 3 & 03302409 + 3114043 & [ edj2009 ] 261 & 17.30 & 6.83e+00 & 7.73e+00 & 6.93e+00 & 4.99e+00 & 3.96e+00 & 3.80e+00 & 5.04e+00 & 5.43e+00 & @xmath3 1.05e+01 & per + 4 & 03335108 + 3112278 & [ edj2009 ] 300 & 16.72 & 4.15e+01 & 6.37e+01 & 5.87e+01 & 3.50e+01 & 2.50e+01 & 1.72e+01 & 1.07e+01 & 5.06e+00 & @xmath3 1.53e+01 & per + 5 & 03344987 + 3115498 & [ edj2009 ] 303 & 13.79 & 8.34e+01 & 1.18e+02 & 1.04e+02 & 6.71e+01 & 4.57e+01 & 3.44e+01 & 3.43e+01 & 4.86e+01 & 4.04e+01 & per + 6 & 03411412 + 3159462 & [ edj2009 ] 307 & 12.59 & 1.57e+02 & 1.37e+02 & 1.06e+02 & 5.00e+01 & 3.24e+01 & 2.34e+01 & 1.85e+01 & 9.64e+00 & @xmath3 6.53e+01 & per + 7 & 03413918 + 3136106 & bd+31 634 & 10.00 & 2.37e+02 & 1.68e+02 & 1.16e+02 & 5.05e+01 & 3.61e+01 & 2.99e+01 & 5.26e+01 & 1.51e+02 & @xmath3 5.75e+01 & per + 8 & 03422333 + 3157426 & [ edj2009 ] 323 & 12.23 & 1.60e+02 & 1.39e+02 & 1.07e+02 & 5.22e+01 & 3.39e+01 & 2.50e+01 & 2.19e+01 & 9.45e+00 & @xmath3 2.61e+01 & per + 9 & 03434461 + 3208177 & cl * ic 348 lrl 67 & 14.65 & 2.41e+01 & 3.61e+01 & 3.21e+01 & 1.90e+01 & 1.49e+01 & 1.14e+01 & 1.09e+01 & 1.04e+02 & 1.73e+02 & per + 10 & 03440915 + 3207093 & cl * ic 348 lrl 8 & 9.41 & 4.55e+02 & 3.24e+02 & 2.39e+02 & 9.79e+01 & 6.64e+01 & 4.63e+01 & 3.68e+01 & 1.00e+01 & @xmath3 1.44e+02 & per + 11 & 03441912 + 3209313 & cl * ic 348 lrl 30 & 11.24 & 1.37e+02 & 1.18e+02 & 8.32e+01 & 3.24e+01 & 2.33e+01 & 1.60e+01 & 9.81e+00 & 3.78e+00 & @xmath3 3.02e+02 & per + 12 & 03442156 + 3215098 & cl * ic 348 lrl 185 & & 1.01e+01 & 1.31e+01 & 1.15e+01 & 5.66e+00 & 4.29e+00 & 3.04e+00 & 1.99e+00 & 4.84e+00 & @xmath3 1.71e+02 & per + 13 & 03442257 + 3201536 & cl * ic 348 lrl 72 & 15.93 & 2.26e+01 & 3.55e+01 & 3.23e+01 & 1.72e+01 & 1.11e+01 & 8.91e+00 & 7.55e+00 & 6.76e+01 & @xmath3 7.36e+02 & per + 14 & 03443200 + 3211439 & cl * ic 348 lrl 12b & 12.88 & 1.49e+02 & 1.84e+02 & 1.80e+02 & 1.38e+02 & 1.08e+02 & 2.61e+02 & 7.66e+02 & 3.68e+02 & @xmath3 8.65e+02 & per + 15 & 03443694 + 3206453 & cl * ic 348 lrl 6 & 11.69 & 3.31e+02 & 4.13e+02 & 3.52e+02 & 2.08e+02 & 1.48e+02 & 1.16e+02 & 9.26e+01 & 4.61e+01 & @xmath3 5.38e+02 & per + 16 & 03444351 + 3207427 & cl * ic 348 lrl 52 & 17.01 & 2.27e+01 & 4.53e+01 & 4.54e+01 & 2.68e+01 & 1.83e+01 & 1.26e+01 & 8.65e+00 & 9.77e+00 & @xmath3 6.80e+02 & per + 17 & 03445614 + 3209152 & cl * ic 348 lrl 21 & 14.81 & 6.21e+01 & 1.03e+02 & 1.08e+02 & 8.69e+01 & 6.53e+01 & 5.53e+01 & 4.38e+01 & 2.14e+02 & @xmath3 2.30e+02 & per + 18 & 03450142 + 3205017 & cl * ic 348 lrl 25 & 12.38 & 1.19e+02 & 1.13e+02 & 9.24e+01 & 4.36e+01 & 3.01e+01 & 2.04e+01 & 1.36e+01 & 4.44e+00 & @xmath3 8.10e+01 & per + 19 & 04104210 + 3805598 & bd+37 887 & 8.47 & 1.41e+03 & 1.10e+03 & 7.76e+02 & 3.41e+02 & 2.10e+02 & 1.51e+02 & 8.71e+01 & 4.39e+01 & @xmath3 5.77e+01 & aur + 20 & 04190110 + 2819420 & [ gba2007 ] 527 & 16.50 & 9.81e+01 & 1.48e+02 & 1.49e+02 & 8.79e+01 & 6.12e+01 & 4.42e+01 & 3.19e+01 & 2.43e+02 & 4.10e+02 & tau + 21 & 04192625 + 2826142 & v819 tau & 12.01 & 2.52e+02 & 3.56e+02 & 2.85e+02 & 1.47e+02 & 8.64e+01 & 6.58e+01 & 3.82e+01 & 2.11e+01 & @xmath3 3.35e+01 & tau + 22 & 04210934 + 2750368 & & 15.55 & 5.14e+01 & 5.59e+01 & 4.79e+01 & 2.90e+01 & 2.19e+01 & 1.66e+01 & 1.32e+01 & 9.46e+00 & @xmath3 2.25e+01 & tau + 23 & 04242321 + 2650084 & & 14.48 & 7.57e+01 & 1.00e+02 & 8.94e+01 & 4.61e+01 & 3.13e+01 & 2.24e+01 & 1.33e+01 & 4.95e+00 & @xmath3 3.23e+01 & tau + 24 & 04284263 + 2714039 & wds j04287 + 2714ab & 16.81 & 2.29e+01 & 3.83e+01 & 4.39e+01 & 3.50e+01 & 2.76e+01 & 2.39e+01 & 1.89e+01 & 2.27e+01 & @xmath3 2.29e+01 & tau + 25 & 04292083 + 2742074 & iras 04262 + 2735 & 15.54 & 6.06e+02 & 1.06e+03 & 1.04e+03 & 6.04e+02 & 4.17e+02 & 3.69e+02 & 5.95e+02 & 4.28e+02 & @xmath3 2.63e+01 & tau + 26 & 04292971 + 2616532 & fw tau & 14.65 & 1.16e+02 & 1.38e+02 & 1.17e+02 & 6.48e+01 & 4.48e+01 & 3.24e+01 & 1.80e+01 & 6.79e+00 & @xmath3 2.35e+01 & tau + 27 & 04295531 + 2258579 & iras 04269 + 2252 & 13.13 & 4.41e+03 & 8.11e+03 & 8.64e+03 & 2.70e+03 & 1.93e+03 & 1.82e+03 & 1.45e+03 & 5.17e+02 & @xmath3 1.94e+01 & tau + 28 & 04300113 + 3517247 & hbc 390 & 15.51 & 7.47e+01 & 1.37e+02 & 1.30e+02 & 7.57e+01 & 4.80e+01 & 3.08e+01 & 2.29e+01 & 2.65e+01 & @xmath3 1.43e+03 & aur + 29 & 04300424 + 3522238 & & 16.53 & 8.21e+00 & 1.30e+01 & 1.13e+01 & 6.47e+00 & 4.25e+00 & 3.23e+00 & 2.69e+00 & 2.48e+01 & @xmath3 3.54e+02 & aur + 30 & 04301644 + 3525217 & [ had2004 ] ldn 1482 g & 12.58 & 9.60e+01 & 8.15e+01 & 6.20e+01 & 2.78e+01 & 1.93e+01 & 1.33e+01 & 1.00e+01 & 6.63e+00 & @xmath3 9.85e+01 & aur + 31 & 04303235 + 3536133 & & 15.94 & 3.13e+01 & 6.07e+01 & 6.08e+01 & 3.37e+01 & 2.42e+01 & 1.72e+01 & 1.55e+01 & 2.92e+02 & @xmath3 9.60e+01 & aur + 32 & 04304004 + 3542101 & & 15.76 & 1.83e+01 & 2.73e+01 & 2.49e+01 & 1.44e+01 & 1.02e+01 & 7.32e+00 & 5.57e+00 & 7.36e+00 & @xmath3 4.46e+01 & aur + 33 & 04305137 + 2442222 & zz tau & 13.31 & 2.54e+02 & 3.41e+02 & 2.80e+02 & 1.65e+02 & 1.24e+02 & 1.00e+02 & 1.09e+02 & 1.07e+02 & @xmath3 9.50e+01 & tau + 34 & 04312113 + 2658422 & iras 04282 + 2652 & 14.76 & 3.14e+03 & 6.25e+03 & 7.24e+03 & 2.46e+03 & 1.92e+03 & 2.03e+03 & 1.29e+03 & 3.84e+02 & @xmath3 3.00e+01 & tau + 35 & 04314503 + 2859081 & & 13.32 & 4.14e+01 & 3.52e+01 & 2.68e+01 & 1.16e+01 & 7.45e+00 & 4.89e+00 & 3.03e+00 & 2.18e+00 & @xmath3 1.54e+01 & tau + 36 & 04330422 + 2921499 & bd+29 719 & 9.96 & 7.02e+02 & 5.50e+02 & 3.93e+02 & 1.79e+02 & 1.22e+02 & 8.49e+01 & 8.72e+01 & 4.31e+02 & @xmath3 3.22e+02 & tau + 37 & 04343549 + 2644062 & & 13.90 & 5.22e+02 & 1.11e+03 & 1.17e+03 & 6.57e+02 & 3.97e+02 & 3.02e+02 & 2.09e+02 & 8.01e+01 & @xmath3 1.28e+01 & tau + 38 & 04364912 + 2412588 & hd 283759 & 10.01 & 5.27e+02 & 4.32e+02 & 3.09e+02 & 1.12e+02 & 8.25e+01 & 5.50e+01 & 3.36e+01 & 1.53e+01 & 2.80e+02 & tau + 39 & 04372486 + 2709195 & hd 283751 & 11.25 & 1.94e+02 & 1.55e+02 & 1.27e+02 & 4.65e+01 & 3.41e+01 & 2.70e+01 & 2.07e+01 & 8.21e+00 & @xmath3 2.37e+01 & tau + 40 & 04385827 + 2631084 & elia 3 - 14 & 15.60 & 2.73e+02 & 8.58e+02 & 1.16e+03 & 7.48e+02 & 4.85e+02 & 3.82e+02 & 2.42e+02 & 9.58e+01 & @xmath3 3.24e+01 & tau + 41 & 04403979 + 2519061 & wds j04407 + 2519ab & 17.87 & 2.98e+01 & 4.94e+01 & 5.36e+01 & 3.27e+01 & 2.41e+01 & 1.63e+01 & 9.49e+00 & 6.64e+00 & @xmath3 3.41e+01 & tau rrrlrrrrrrrr [ observed ] 1 & 52.3617 & 31.4465 & m2 & 140 & 1.30 & @xmath3 2.9 & & 0.25 ; 1.7 & 188 ; 63 & 0.83 ; 2.09 + 2 & 52.3720 & 31.3096 & m0 & 280 & 1.30 & 6.3 & 1.1 & & & @xmath13 1.65 ; 3.85 + 3 & 52.6003 & 31.2345 & m2 & 340 & 1.30 & @xmath3 3.0 & & & & @xmath13 1.82 ; 3.70 + 4 & 53.4628 & 31.2077 & m3 & 170 & 1.30 & @xmath3 2.9 & & 0.96 & 101 & 2.08 + 5 & 53.7077 & 31.2640 & m2 & 450 & 1.30 & 6.0 & 1.2 & 0.85 & 231 & 2.62 + 6 & 55.3088 & 31.9962 & a3 & -1 & 0.85 & @xmath3 15.0 & & & & @xmath13 2.95 ; 5.23 + 7 & 55.4132 & 31.6030 & a5 & -1 & 1.30 & @xmath3 3.0 & & & & @xmath13 2.62 ; 4.86 + 8 & 55.5972 & 31.9619 & a0 & -1 & 0.85 & @xmath3 15.0 & & & & @xmath13 3.08 ; 4.73 + 9 & 55.9359 & 32.1383 & m1 & 280 & 0.85 & 31 & 6.0 & & & @xmath13 1.91 ; 4.01 + 10 & 56.0382 & 32.1192 & a2 & -1 & & & & & & @xmath13 2.64 ; 4.33 + 11 & 56.0797 & 32.1587 & f0 & -1 & & & & & & @xmath13 2.77 ; 4.49 + 12 & 56.0899 & 32.2527 & m2 & 130 & 1.30 & @xmath3 3.0 & & & & @xmath13 0.95 ; 3.14 + 13 & 56.0941 & 32.0316 & m2 & 140 & 1.30 & @xmath3 1.5 & & & & @xmath13 1.23 ; 3.03 + 14 & 56.1335 & 32.1955 & a5 & -1 & 1.30 & @xmath3 2.8 & & 1.31 & 274 & 0.20 + 15 & 56.1540 & 32.1126 & g2 & -1 & 1.30 & @xmath3 2.9 & & 0.56 & 207 & 1.12 + 16 & 56.1814 & 32.1285 & m1 & 150 & 1.30 & @xmath3 2.5 & & & & @xmath13 1.67 ; 2.25 + 17 & 56.2339 & 32.1542 & k0 & 360 & 1.30 & @xmath3 1.1 & & & & @xmath13 2.15 ; 3.61 + 18 & 56.2559 & 32.0838 & a3 & -1 & & & & 0.65 & 271 & 2.89 + 19 & 62.6755 & 38.1000 & a3 & -1 & 1.30 & @xmath3 3.9 & & & & @xmath13 2.65 ; 4.59 + 20 & 64.7546 & 28.3284 & m6 & 210 & 1.30 & @xmath3 3.3 & & & & @xmath13 1.94 ; 3.94 + 21 & 64.8594 & 28.4373 & k7 & 180 & 1.30 & @xmath3 5.4 & & & & @xmath13 2.74 ; 4.03 + 22 & 65.2889 & 27.8436 & m4 & 280 & 1.30 & @xmath3 3.0 & & 0.77 & 230 & 1.28 + 23 & 66.0967 & 26.8357 & m2 & 200 & & & & & & @xmath13 1.93 ; 3.19 + 24 & 67.1776 & 27.2344 & m4 & 300 & 1.30 & @xmath3 3.0 & & 0.64 & 10 & 0.94 + 25 & 67.3368 & 27.7021 & agb & -1 & & & & & & @xmath13 2.27 : 3.72 + 26 & 67.3738 & 26.2814 & m5 & 190 & 0.85 & 4.5 & 1.1 & 0.05 & 45 & @xmath600 + 27 & 67.4805 & 22.9828 & agb & -1 & 1.30 & @xmath3 6.0 & & & & @xmath13 3.23 ; 3.99 + 28 & 67.5048 & 35.2902 & k7 & 370 & 1.30 & @xmath3 3.3 & & 1.3 & 24 & 1.01 + 29 & 67.5177 & 35.3733 & m0 & 370 & 1.30 & 9.7 & 1.5 & & & @xmath13 2.12 ; 3.67 + 30 & 67.5685 & 35.4227 & a4 & -1 & 1.30 & @xmath3 3.3 & & & & @xmath13 2.41 ; 4.29 + 31 & 67.6348 & 35.6037 & k7 & 350 & 0.85 & 10.0 & 2.0 & 0.83 & 61 & 2.30 + 32 & 67.6669 & 35.7029 & k7 & 310 & 1.30 & @xmath3 2.5 & & 1.2 & 307 & @xmath0 1 + 33 & 67.7141 & 24.7062 & m3 & 330 & 0.85 & @xmath3 8.0 & & 0.07 & 315 & @xmath0 0 + 34 & 67.8381 & 26.9784 & agb & -1 & 1.30 & @xmath3 5.0 & & & & @xmath13 1.69 ; 2.66 + 35 & 67.9377 & 28.9856 & f5 & -1 & & & & & & @xmath13 3.41 ; 4.94 + 36 & 68.2676 & 29.3639 & b9 & -1 & 1.30 & @xmath3 2.8 & & & & @xmath13 2.07 ; 3.88 + 37 & 68.6479 & 26.7351 & agb & -1 & 0.85 & @xmath3 30.0 & & & & @xmath13 2.06 ; 4.33 + 38 & 69.2047 & 24.2163 & f2 & -1 & 0.85 & @xmath3 27.0 & & & & @xmath13 2.73 ; 3.96 + 39 & 69.3536 & 27.1554 & agb & -1 & 1.30 & @xmath3 3.3 & & & & @xmath13 3.11 ; 5.41 + 40 & 69.7428 & 26.5190 & agb & -1 & 0.85 & @xmath3 30.0 & & & & @xmath13 1.80 ; 3.66 + 41 & 70.1658 & 25.3184 & m5 & 130 & 1.30 & @xmath3 2.9 & & & & @xmath13 1.74 ; 2.34 lrrrrrrrc [ derived ] 1 & @xmath3 -11.0 & @xmath3 2.6 & 62 ; 425 & 5.8 & -1.19 & -2.89 & 0.6 & photoevaporating disk + 2 & -10.2 & 5.6 & & 5.8 & 0.69 & -2.33 & 1.3 & giant planet - forming disk + 3 & -9.6 & @xmath3 2.7 & & 3.6 & -0.95 & -1.72 & 0.1 & grain growth - dominated disk + 4 & @xmath3 -11.0 & @xmath3 2.6 & 240 & 8.0 & -1.69 & @xmath3 -3.57 & 0.5 & debris disk + 5 & -8.5 & 5.3 & 212 & 5.8 & -0.80 & -2.56 & 0.5 & grain growth - dominated disk + 6 & @xmath3 -11.0 & @xmath3 6.2 & & 5.8 & -1.68 & -4.27 & 0.8 & debris disk + 7 & @xmath3 -11.0 & @xmath3 2.7 & & 5.8 & 0.11 & -3.09 & 0.1 & photoevaporating disk + 8 & @xmath3 -11.0 & @xmath36.2 & & 5.8 & -1.74 & -4.09 & 0.8 & debris disk + 9 & -10.2 & 13.0 & & 5.8 & 0.49 & -2.19 & 0.7 & giant planet - forming disk + 10 & -10.9 & & & 5.8 & -2.12 & -4.39 & 0.4 & debris disk + 11 & @xmath3 -11.0 & & & 8.0 & -1.87 & @xmath3 -3.40 & 0.1 & debris disk + 12 & @xmath3 -11.0 & @xmath3 2.70 & & 8.0 & -0.20 & @xmath3 -2.41 & 0.5 & photoevaporating disk + 13 & @xmath3 -11.0 & @xmath3 1.3 & & 5.8 & 0.33 & -2.92 & 1.5 & photoevaporating disk + 14 & @xmath3 -11.0 & @xmath3 2.5 & 327 & 3.6 & -0.76 & -2.37 & 0.5 & photoevaporating disk + 15 & @xmath3 -11.0 & @xmath3 2.6 & 140 & 4.5 & -1.75 & -2.80 & 1.0 & photoevaporating disk + 16 & @xmath3 -11.0 & @xmath3 2.2 & & 8.0 & -0.95 & @xmath3 -3.03 & 2.0 & debris disk + 17 & -9.4 & @xmath3 1.0 & & 4.5 & -0.35 & -2.26 & 1.2 & giant planet - forming disk + 18 & @xmath3 -11.0 & & 162 & 8.0 & -2.03 & @xmath3 -4.05 & 0.5 & debris disk + 19 & @xmath3 -11.0 & @xmath3 3.0 & & 8.0 & -1.63 & @xmath3 -4.88 & 0.4 & debris disk + 20 & -10.85 & @xmath3 0.6 & & 8.0 & 0.83 & -2.14 & 0.6 & giant planet - forming disk + 21 & @xmath3 -11.0 & @xmath3 0.9 & & 8.0 & -1.55 & @xmath3 -3.90 & 0.5 & debris disk + 22 & -10.3 & @xmath3 0.50 & 108 & 5.8 & -1.42 & -2.57 & 0.0 & grain growth - dominated disk + 23 & @xmath3 -11.0 & & & 8.0 & -1.91 & @xmath3 -3.78 & 0.4 & debris disk + 24 & -10.0 & @xmath3 0.50 & 90 & 4.5 & -1.18 & -2.33 & 1.2 & grain growth - dominated disk + 25 & & & & 5.8 & -0.95 & & & agb star + 26 & @xmath3 -11.0 & 0.4 & 7 & 8.0 & -1.88 & @xmath3 -3.77 & 0.0 & photoevaporating / circumbinary disk + 27 & @xmath3 -11.0 & @xmath3 1.0 & & 8.0 & -1.97 & & & agb star + 28 & -9.3 & @xmath3 2.6 & 390 & 8.0 & -0.90 & @xmath3 -2.53 & 1.1 & grain growth - dominated disk + 29 & -9.3 & 7.5 & & 5.8 & 0.38 & -2.62 & 0.7 & giant planet - forming disk + 30 & @xmath3 -11.0 & @xmath3 2.6 & & 8.0 & -1.39 & @xmath3 -4.28 & 0.7 & ms debris disk + 31 & -9.5 & 3.7 & 249 & 5.8 & 0.90 & -2.15 & 1.3 & giant planet - forming disk + 32 & -9.9 & @xmath3 1.9 & 360 & 5.8 & -1.05 & -2.73 & 0.6 & grain growth - dominated disk + 33 & -9.7 & @xmath3 2.9 & 10 & 5.8 & -0.99 & -2.37 & 0.3 & grain growth - dominated disk + 34 & & @xmath3 0.8 & & 3.6 & -0.99 & & & agb star + 35 & @xmath3 -11.0 & & & 8.0 & -1.32 & @xmath3 -3.65 & 0.7 & ms debris disk + 36 & @xmath3 -11.0 & @xmath3 0.5 & & 5.8 & 0.09 & -3.44 & 0.5 & ms debris disk + 37 & & & & 8.0 & -1.94 & & & agb star + 38 & @xmath3 -11.0 & & & 8.0 & -1.71 & @xmath3 -3.15 & 0.0 & ms debris disk + 39 & & @xmath3 0.6 & & 8.0 & -1.86 & & & agb star + 40 & & & & 8.0 & -1.93 & & & agb star + 41 & @xmath3 -11.0 & @xmath3 0.5 & & 8.0 & -1.35 & @xmath3 -3.90 & 0.9 & debris disk lcccc [ incidence ] all stars & 39.2@xmath17 ( 29/74 ) & 17.6@xmath17 ( 13/74 ) & 20.3@xmath17 ( 15/74 ) & 23.0@xmath17 ( 17/74 ) + m - type stars & 39.6@xmath17 ( 19/48 ) & 16.6@xmath17 ( 8/48 ) & 25.0@xmath17 ( 12/48 ) & 18.8@xmath17 ( 9/48 ) + k - type stars & 58.8@xmath17 ( 10/17 ) & 29.4@xmath17(5/17 ) & 0.0@xmath17 ( 0/17 ) & 11.8@xmath17 ( 2/17 ) + gafb - type stars & 0.0@xmath17 ( 0/9 ) & 0.0@xmath17 ( 0/9 ) & 33.3@xmath17 ( 3/9 ) & 66.6@xmath17 ( 6/9 ) + rcccc [ incidence ] @xmath3 1 myr & 71.4@xmath17 ( 10/14 ) & 0.0@xmath17 ( 0/14 ) & 14.3@xmath17 ( 2/14 ) & 14.3@xmath17 ( 2/14 ) + 1 - 3 myr & 33.3@xmath17 ( 10/30 ) & 26.7@xmath17 ( 8/30 ) & 20.0@xmath17 ( 6/30 ) & 20.0@xmath17 ( 6/30 ) + @xmath13 3 myr & 30.0@xmath17 ( 9/30 ) & 16.7@xmath17 ( 5/30 ) & 23.3@xmath17 ( 7/30 ) & 30.0@xmath17 ( 9/30 ) + -[4.5 ] @xmath3 0.25 and [ 3.6 ] -[24 ] @xmath3 0.5 are consistent with bare stellar photospheres . blue dots are disk - less wttss from cieza et al .
( 2007 ) used to define this region of the diagram .
red stars and triangles are all the young stellar objects candidates ( ysoc ) in the perseus and auriga catalogs from the _ c2d _ and _ gould belt _ legacy projects . since the catalogs from the _ taurus _ project does not provide a ysoc classification , we simply plot all the targets with ir excesses and 3.6 and 24 @xmath14 m fluxes at least as large as those of the faintest transition disks in our sample ( red diamonds ) .
most pms stars are either disk - less or have excesses at both 4.5 and 24 @xmath14 m . our 31 transition disks , shown as green dots , have significant ( @xmath13 5-@xmath15 ) excess at 24 @xmath14 m and little or no excess at 4.5 @xmath14 m , as expected for disks with inner holes . , width=432 ] r diagram diagram . the 0.5
, 1 , 3 , and 10 myr isochrones from the siess et al .
( 2000 ) models are also shown .
five of the six agb stars are clearly overluminous ( targets # 25 , 27 , 34 , 37 , and 40 ) . three stars in taurus ( # 35 , 36 , and 38 ) and one in auriga ( # 30 ) are significantly underluminous and are likely to be background ms stars .
one target in perseus ( # 11 ) is also slightly underluminous , but it has been classified as a member of the ic 348 cluster based on proper motion measurements ( luhman et al .
width=576 ] velocity profiles of the 13 accreting objects in our sample .
the dashed line indicates the 10@xmath17 peak intensity , where @xmath61v is measured .
the intervals delimited by the dotted and dashed - dotted lines correspond to @xmath61v = 200 and @xmath61v = 300 km / s , respectively .
all accreting objects have @xmath61v @xmath13 200 km / s . , width=528 ] velocity profiles of the 9 non - accreting objects in our sample where h@xmath25 was detected in emission .
they are all k and m - type stars .
the dashed line indicates the 10@xmath17 peak intensity , where @xmath61v is measured .
the intervals delimited by the dotted and dashed - dotted lines correspond to @xmath61v = 200 and @xmath61v = 300 km / s , respectively .
non - accreting objects show symmetric and narrow ( @xmath61v @xmath10 200 km / s ) h@xmath25 emission , consistent with chromospheric activity .
, width=528 ] limits .
the open squares correspond to the observed optical and near - ir fluxes before being corrected for extinction using the a@xmath62 values listed in table 3 ( calculated as described in section [ pms_id ] ) and the extinction curve provided by the asiago database of photometric systems ( fiorucci @xmath1 munari 2003 ) .
the solid green lines represent the stellar photosphere normalized to the extinction - corrected j - band .
the dotted lines correspond to the median mid - ir sed of k5m2 cttss calculated by furlan et al .
the dashed lines are the quartiles .
, width=480 ] 5 m@xmath4 ) , and/or fractional disk luminosities l@xmath63/l@xmath64 .
as discussed in section [ pms_id ] , objects # 30 , 35 , 36 , and 38 could be debris disks around early - type background ms stars .
, width=624 ] mayor , 1991 ) is shown for comparison .
binary systems with separations in the 10 to 30 au range result in the rapid erosion of the individual circumstellar disks .
few circumbinary disks survive in such systems .
, width=528 ] .
the lack of non - accreting objects with relative massive outer disks ( @xmath24 25 m@xmath4 ) favors photoevaporation models with low evaporation rates ( @xmath010@xmath55m@xmath27yr@xmath53 ) across a wide range of stellar masses . ,
width=672 ] , but showing the location of each type of disk for the combined sample of disks from papers i , ii , and iii .
all stars hotter than @xmath010@xmath58 ( 5754 k , corresponding to a g5 star ) have non - accreting disks , either photoevaporating disks or debris disk , consistent with the idea that primordial disks dissipate faster around more massive objects .
there is a lack of ( giant ) planet - forming disk candidates among the youngest stars in the sample .
this favors core accretion as the main planet formation mechanism and a 2 to 3 myr formation timescale .
, width=672 ] | as part of an ongoing program aiming to characterize a large number of _ spitzer_-selected transition disks ( disks with reduced levels of near - ir and/or mid - ir excess emission ) , we have obtained ( sub)millimeter wavelength photometry , high - resolution optical spectroscopy , and adaptive optics near - infrared imaging for a sample of 31 transition objects located in the perseus , taurus , and auriga molecular clouds .
we use these ground - based data to estimate disk masses , multiplicity , and accretion rates in order to investigate the mechanisms potentially responsible for their inner holes .
following our previous studies in other regions , we combine disk masses , accretion rates and multiplicity data with other information , such as sed morphology and fractional disk luminosity to classify the disks as _
strong candidates _ for the following categories : grain - growth dominated disks ( 7 objects ) , giant planet - forming disks ( 6 objects ) , photoevaporating disks ( 7 objects ) , debris disks ( 11 objects ) , and cicumbinary disks ( 1 object , which was also classified as a photoeavaporating disk ) . combining our sample of 31 transition disks with those from our previous studies results in a sample of 74 transition objects that have been selected , characterized , and classified in an homogenous way .
we discuss this combined high - quality sample in the context of the current paradigm of the evolution and dissipation of protoplanetary disks and use its properties to constrain different aspects of the key processes driving their evolution .
we find that the age distribution of disks that are likely to harbor recently formed giant planets favors core accretion as the main planet formation mechanism and a @xmath02 - 3 myr formation timescale . |
with the rapid growth of wireless data traffic , multi - user interference has become a major bottleneck that limits the performance of the existing wireless communication services .
mathematically , we can model such an interference - limited communication system as a multi - user interference channel in which a number of linearly interfering transmitters simultaneously send private data to their respective receivers . exploiting time / space / frequency / code diversity are effective approaches to mitigate / manage multi - user interference .
for instance , when the transmitters and/or receivers are equipped with multiple antennas , a joint optimization of the physical layer transmit - receive beamforming vectors for all users can efficiently mitigate multi - user interference@xcite ; when all transmitters and receivers are equipped with a single antenna , one way to control / mitigate multi - user interference is to impose certain frequency restrictions or transmission power limits on the frequency resources used by each transmitter@xcite . in particular , orthogonal frequency division multiple access ( ofdma ) is a form of multi - carrier transmission and is suited for frequency selective channels and high data rates .
this technique effectively decomposes a frequency - selective wide - band channel into a group of non - selective narrowband subchannels ( subcarriers ) , which makes it robust against large delay spreads by preserving orthogonality in the frequency domain .
moreover , the ingenious introduction of cyclic redundancy at the transmitter reduces the complexity to only fft processing and one tap scalar equalization at the receiver@xcite . conventional ofdma schemes _ preassign _ subcarriers to users in a nonoverlapping way , thus users ( transmitting on different subcarriers ) cause no interference to each other .
although the ofdma scheme is well suited to be used in a high - speed communication context where quality of service is a major concern , it can lead to inefficient bandwith utilization .
this is because that the preassignment of subcarries can not adapt traffic load and channel fluctuations in space and time ; i.e. , a subcarrier preassigned to a user can not be released to other users even if it is unusable when the user s channel conditions are poor . to adapt these fluctuations and improve the overall system s throughput , ofdma based subcarrier allocation networks such as worldwide interoperability for microwave access ( wimax)@xcite and long term evolution ( lte)@xcite should be equipped with _
dynamic _ subcarrier and power allocation algorithms .
in particular , a dynamic ofdma based subcarrier and power allocation algorithm is well suited for the _ dense _ femtocell downlink system@xcite , where a large number of femtocells close to each other are deployed in a macrocell .
the joint optimization of subcarrier and power allocations for the multi - user ofdma system is a nonconvex problem , therefore various heuristics approaches have been proposed for this problem @xcite .
very recently , the authors in @xcite proposed a dynamic joint frequency and transmission power allocation scheme called enhanced dynamic frequency planning ( edfp ) .
edfp is a hybrid centralized and distributed architecture , where a central broker first dynamically partitions subcarriers among neighboring cells so that the long - term cell - edge inter - cell interference is minimized , and then each cell independently allocates subcarriers and transmission power to its users in a way that its total transmission power is minimized .
notice that nonconvex optimization problems are difficult to solve in general .
however , not all nonconvex problems are hard to handle since the lack of convexity may be due to an inappropriate formulation , and many nonconvex optimization problems indeed admit a convex reformulation @xcite
. therefore , nonconvexity of the joint subcarrier and power allocation problem for the multi - user ofdma system does not imply that it is computationally intractable ( strongly np - hard in terms of computational complexity theory@xcite ) .
the aim of this paper is to characterize the computational complexity of the joint subcarrier and power allocation problem for the multi - user ofdma system ; i.e. , to categorize when the problem is ( strongly ) np - hard and when it is polynomial time solvable .
in fact , the dynamic spectrum management problem _ without the ofdma constraint
_ ( at most one user is allowed to transmit power on each subcarrier ) has been extensively studied in @xcite .
it is shown in @xcite that the dynamic spectrum management problem is ( strongly ) np - hard when the number of subcarriers is greater than two , or when the number of users is greater than one .
however , the analysis of these results is highly dependent on the crosstalk channel gains among users ; i.e. , some of the crosstalk channel gains are assumed to be large enough , and some of them are assumed to be zero .
we shall see late ( in section [ sec - model ] ) that the magnitude of crosstalk channel gains has no influence on the user s transmission rate if all users are required to transmit power in an orthogonal manner .
this makes the two problems ( the dynamic spectrum management problem without the ofdma constraint in @xcite and the joint subcarrier and power allocation problem with the ofdma constraint considered in this paper ) sharply different from each other .
an interesting result shown in @xcite is that the optimal solution of the two - user _ sum - rate _ maximization problem is automatically ofdma if the crosstalk channel gains of the two users on each subcarrier are large enough .
in addition , based on the fact that the duality gap of the dynamic spectrum management problem is zero as the number of subcarriers goes to infinity ( regardless of the convexity of the objective function)@xcite , dual decomposition algorithms have been proposed in @xcite for the nonconvex optimization problem of multi - carrier systems , again without considering the ofdma constraint . in this paper , we focus on the characterization of the computational complexity status of the joint subcarrier and power allocation problem for the _ multi - user ofdma system_. in particular , we consider two formulations of the joint subcarrier and power allocation problem .
the first one is the problem of minimizing the total transmission power in the system subject to all users quality of service constraints , all users power budget constraints per subcarrier , and the ofdma constraint .
the second one is the problem of maximizing the system utility ( including the sum - rate utility , the proportional fairness utility , the harmonic mean utility , and the min - rate utility ) while the total transmission power constraint of each user , individual power constraints on each subcarrier , and the ofdma constraint are respected .
the main contributions of this paper are twofold .
first , we show that the aforementioned two formulations of the joint subcarrier and power allocation problem are strongly np - hard . the proof is based on a polynomial time transformation from the @xmath0-dimensional matching problem .
the strong np - hardness results suggest that for a given ofdma system , computing the optimal subcarrier and power allocation strategy is generally intractable .
thus , instead of insisting on finding an efficient algorithm that can find the global optimum of the joint subcarrier and power allocation problem , one has to settle with less ambitious goals , such as finding high quality approximate solutions or locally optimal solutions of the problem in polynomial time .
second , we also identify several subclasses of the joint allocation problem which can be solved to global optimality or @xmath1-global optimality in polynomial time .
we therefore clearly delineate the set of computationally tractable problems within the general class of np - hard joint subcarrier and power allocation problems .
specifically , we show in this paper that , when there is only a single user in the system or when the number of subcarriers and the number of users are equal to each other , the total transmission power minimization problem is polynomial time solvable ; when there is only a single user , the aforementioned four utility maximization problems are all polynomial time solvable .
the rest of this paper is organized as follows . in section [ sec - model ] , we introduce the system model and give the two formulations of the joint subcarrier and power allocation problem for the multi - user ofdma system . in section
[ sec - hard ] , we first give a brief introduction to computational complexity theory@xcite and then address the computational complexity of the joint subcarrier and power allocation problem .
in particular , we show that the aforementioned two formulations of the joint subcarrier and power allocation problem are generally strongly np - hard .
several subclasses of the joint allocation problem which are polynomial time solvable are identified in section [ sec - easy ] .
finally , the conclusion is drawn in section [ sec - conclusion ] .
in this section , we introduce the system model and problem formulation . consider a multi - user ofdma system , where there are @xmath3 users ( transmitter - receiver pairs ) sharing @xmath4 subcarriers . throughout the paper , we assume that @xmath5 ; i.e.
, the number of subcarriers is greater than or equal to the number of users . otherwise , the ofdma constraint is infeasible .
denote the set of users and the set of subcarriers by @xmath6 and @xmath7 , respectively .
for any @xmath8 and @xmath9 , suppose @xmath10 to be the symbol that transmitter @xmath11 wishes to send to receiver @xmath11 on subcarrier @xmath12 , then the received signal @xmath13 at receiver @xmath11 on subcarrier @xmath12 can be expressed by @xmath14 where @xmath15 is the channel coefficient from transmitter @xmath16 to receiver @xmath11 on subcarrier @xmath12 and @xmath17 is the additive white gaussian noise ( awgn ) with distribution @xmath18@xmath19 denoting the power of @xmath20 by @xmath21 ; i.e. , @xmath22 , the received power at receiver @xmath11 on subcarrier @xmath12 is given by @xmath23 where @xmath24 stands for the channel gain from transmitter @xmath16 to receiver @xmath11 on subcarrier @xmath25 treating interference as noise , we can write the sinr of receiver @xmath11 on subcarrier @xmath12 as @xmath26 and transmitter @xmath11 s achievable data rate @xmath27 ( bits / sec ) as @xmath28 in this paper , we consider the joint subcarrier and power allocation problem for the multi - user ofdma system .
mathematically , a power allocation vector @xmath29 is said to satisfy the ofdma property if the following equations hold true : @xmath30 the above equations basically say that at most one user is allowed to transmit power on each subcarrier .
therefore , the joint subcarrier and power allocation problem for the multi - user ofdma system can be formulated as @xmath31\text{s.t .
} & r_{k}\geq \gamma_{k},~k\in{{\cal k } } , \\[3pt ] & p_k^n\geq p_k^n\geq 0,~k\in{{\cal k}},~n\in{{\cal n}},\\[3pt ] &
p_k^np_j^n=0,~\forall~j\neq k,{~k,\,j\in{{\cal k}},}~n\in{{\cal n } } , \end{array}\ ] ] where the objective function is the total transmission power of all users on all subcarriers , @xmath32 ( @xmath8 ) is the desired transmission rate target of user @xmath33 @xmath34 is the transmission power budget of user @xmath11 on subcarrier @xmath12 , and the last constraint is the ofdma constraint . due to the ofdma constraint , we know for any @xmath8 and @xmath9 @xmath35 in fact , if @xmath36 , the above equality holds trivially ; while if @xmath37 it follows from the ofdma constraint that @xmath38 and thus @xmath39 which shows that the above equality holds as well .
thus , problem is equivalent to @xmath40\text{s.t . } & \displaystyle\sum_{n\in{{\cal n}}}\log_2\left(1+{\dfrac}{\alpha_{k , k}^np_k^{n}}{\eta_k^n}\right)\geq \gamma_{k},~k\in{{\cal k } } , \\[12pt ] &
p_k^n\geq p_k^n\geq 0,~k\in{{\cal k}},~n\in{{\cal n}},\\[3pt ] & p_k^np_j^n=0,~\forall~j\neq k,{~k,\,j\in{{\cal k}},}~n\in{{\cal n}}. \end{array}\ ] ] by introducing a group of binary variables @xmath41 problem can be reformulated as @xmath42\text{s.t .
} & \displaystyle \sum_{n\in{{\cal n}}}\log_2\left(1+{\dfrac}{\alpha_{k , k}^np_k^{n}}{\eta_k^n}\right)\geq \gamma_{k},~k\in{{\cal k } } , \\[12pt ] & x_k^np_k^n\geq p_k^n\geq 0,~k\in{{\cal k}},~n\in{{\cal n}},\\[3pt ] & x_k^n\in\left\{0,1\right\},~k\in{{\cal k}},~n\in{{\cal n}},\\[3pt ] & \displaystyle\sum_{k\in{{\cal k } } } x_k^n\leq 1,~n\in{{\cal n } } , \end{array}\ ] ] where the binary variable @xmath43 if user @xmath11 transmits power on subcarrier @xmath44 or @xmath45 otherwise .
the last constraint @xmath46 stands for the ofdma constraint .
problem can be dealt with by a two - stage approach . at the first stage
, we solve the subcarrier allocation problem ; i.e. , determining the binary variables @xmath47 , which is equivalent to partitioning the set of subcarriers @xmath7 into @xmath3 nonoverlapping groups @xmath48 . at the second stage
, we solve the power allocation problem ; i.e. , solving @xmath3 decoupled power allocation problems @xmath49\text{s.t .
} & \displaystyle \sum_{n\in{{\cal n}}_k}\log_2\left(1+{\dfrac}{\alpha_{k , k}^np_k^{n}}{\eta_k^n}\right)\geq \gamma_{k } , \\[14pt ] & p_k^n\geq
p_k^n\geq 0,~n\in { \cal n}_k.\\ \end{array}\ ] ] problem at the second stage is convex , and thus is easy to solve . to sum up
, the joint subcarrier and power allocation problem can be equivalently formulated as , , or .
formulation is intuitive and is easy to understand , whereas formulation is compact and is easy to analyze .
the analysis of this paper is mainly based on .
besides the total transmission power minimization problem , we also consider the utility maximization problem for the multi - user ofdma system , which can be expressed by @xmath50 \text{s.t . } & \displaystyle \sum_{n\in{{\cal n}}}p_k^{n}\leq p_{k},~k\in{{\cal k } } , \\[12pt ] & p_k^n\geq p_k^n\geq 0,~k\in{{\cal k}},~n\in{{\cal n}},\\[3pt ] & p_k^np_j^n=0,~\forall~j\neq k,{~k,\,j\in{{\cal k}},}~n\in{{\cal n } } , \end{array}\]]where @xmath51 denotes the system utility function and @xmath52 is the power budget of transmitter @xmath53 four popular system utility functions are * sum - rate utility : @xmath54 * proportional fairness utility : @xmath55 + * harmonic mean utility : @xmath56 + * min - rate utility : @xmath57 it is simple to see that @xmath58 and the equality holds true if and only if @xmath59
in this section , we show that both the total power minimization problem and the system utility maximization problem are intrinsically intractable ( strongly np - hard in the sense of computational complexity theory @xcite ) , provided that the ratio of the number of subcarriers and the number of users , that is @xmath60 , is equal to any given constant number @xmath61 . to begin with ,
we briefly introduce complexity theory in subsection [ sub - i ]
. then we show problems and are strongly np - hard in subsections [ sub - ii ] and [ sub - iii ] , respectively .
in computational complexity theory@xcite , a problem is said to be np - hard if it is at least as hard as any problem in the class np ( problems that are solvable in nondeterministic polynomial time ) .
the np class includes well known problems like the @xmath0-colorability problem ( which is to check whether the nodes of a given graph can be colored in three colors so that each pair of adjacent nodes are colored differently ) .
np - complete problems are the hardest problems in np in the sense that if any np - complete problem is solvable in polynomial time , then each problem in np is solvable in polynomial time . the @xmath0-colorability problem is np - complete .
a problem is strongly np - hard ( strongly np - complete ) if it is np - hard ( np - complete ) and it can not be solved by a pseudo - polynomial time algorithm . an algorithm that solves a problem is called a _
pseudo - polynomial _ time algorithm if its time complexity function is bounded above by a polynomial function related to both of the length and the numerical values of the given data of the problem .
this is in contrast to the polynomial time algorithm whose time complexity function depends only on the length of the given data of the problem .
the @xmath0-colorability problem is strongly np - complete .
however , not all np - hard ( np - complete ) problems are strongly np - hard ( strongly np - complete ) . for instance
, the partition problem is np - hard but not strongly np - hard .
strongly np - hard or np - hard problems may not be in the class np , but they are at least as hard as any np - complete problem .
it is widely believed that there can not exist a polynomial time algorithm to solve any np - complete , np - hard , or strongly np - hard problem ( unless p@xmath2np ) .
thus , once an optimization problem is shown to be np - hard , we can no longer insist on having an efficient algorithm that can find its global optimum in polynomial time .
instead , we have to settle with less ambitious goals , such as finding high quality approximate solutions or locally optimal solutions of the problem in polynomial time .
the standard way to prove an optimization problem is np - hard is to establish the np - hardness of its corresponding feasibility problem or decision problem .
the latter is the problem to decide if the global minimum of the optimization problem is below a given threshold or not .
the output of a decision problem is either true or false .
the feasibility or decision version of an optimization problem is usually in the class np .
clearly , the feasibility or decision version of an optimization problem is always easier than the optimization problem itself , since the latter further requires finding the global minimum value and the minimizer .
thus , if we show the feasibility or decision version of an optimization problem is np - hard , then the original optimization problem must also be np - hard . in complexity theory , to show a decision problem @xmath62 is np - hard , we usually follow three steps : 1 ) choose a suitable known np - complete decision problem @xmath63 2 ) construct a _ polynomial time _
transformation from any instance of @xmath64 to an instance of @xmath65 3 ) prove under this transformation that any instance of problem @xmath64 is true if and only if the constructed instance of problem @xmath62 is true .
furthermore , if the chosen np - complete problem @xmath64 is strongly np - complete , then problem @xmath62 is strongly np - hard . in the following two subsections ,
we show that problems and are strongly np - hard .
to analyze the computational complexity of problem , we consider its feasibility problem . if the feasibility problem is strongly np - hard , so is the original optimization problem .
* feasibility problem of * . given a set of transmission rate levels @xmath67 individual power budgets per subcarrier @xmath34
, direct - link channel gains @xmath68 and noise powers @xmath69 check whether there exists a subcarrier and power allocation strategy such that @xmath70 & p_k^n\geq p_k^n\geq 0,~k\in{{\cal k}},~n\in{{\cal n}},\\[5pt ] & p_k^np_j^n=0,~\forall~j\neq k,{~k,\,j\in{{\cal k}},}~n\in{{\cal n}}. \end{array}\right.\ ] ] to analyze the computational complexity of the feasibility problem , we choose the following strongly np - complete problem ( see @xcite ) .
* @xmath0-dimensional matching problem with size @xmath3 .
* let @xmath71 and @xmath72 be three different sets with @xmath73 and @xmath74 be a subset of @xmath75 .
the @xmath0-dimensional matching problem is to check whether there exists a match @xmath76 such that the following two conditions are satisfied : * for any two different tripes @xmath77 and @xmath78 we have @xmath79 @xmath80 and @xmath81 * @xmath82 , @xmath83 and @xmath84 next , we first show that any instance of the @xmath0-dimensional matching problem corresponds to an instance of the transmission rate feasibility problem when @xmath85 .
[ basic lemma ] [ thm - feasi]checking the feasibility problem of is strongly np - hard when @xmath85 .
thus , problem itself is also strongly np - hard when @xmath85 .
assume that @xmath85 .
consider any instance of the @xmath0-dimensional matching problem with @xmath86@xmath87@xmath88 and a relationship set @xmath89 we construct a multi - user multi - carrier system where there are @xmath3 users ( which correspond to set @xmath90 ) and @xmath91 subcarriers ( which correspond to set @xmath92 ) .
more exactly , @xmath93 and @xmath94 define @xmath95 & { { \cal s}}_2=\left\{(k_x , l_z)\,|\,\left(k_x , j_y , l_z\right)\in { { \cal r}}\right\}.\\ \end{array}\ ] ] for each @xmath96 , the power budgets per subcarrier @xmath34 , the noise powers @xmath69 and the direct - link channel gains @xmath97 are given by @xmath98 2 , \quad \text{if}~n\in { { { \cal z}}},\\[3pt ] \end{array}\right.\ ] ] @xmath99 2 , \quad \text{if}~(k , n)\in { { \cal s}}_2;\\[3pt ] 3 , \quad \text{if}~(k , n)\notin { { \cal s}}_1\bigcup { { \cal s}}_2 , \end{array}\right.\ ] ] and @xmath100 0.25 , \quad \text{if}~(k , n_1,n_2)\notin { { { \cal r } } } , \end{array}\right.\ ] ] respectively . letting @xmath101 the corresponding instance of problem is @xmath102 & 3\geq p_k^n\geq 0,~k\in{{\cal k}},~n\in{{\cal y}},\\[5pt ] & 2\geq p_k^n\geq 0,~k\in{{\cal k}},~n\in{{\cal z}},\\[5pt ] & p_k^np_j^n=0,~\forall~j\neq k,{~k,\,j\in{{\cal k}},}~n\in{{\cal n } } , \end{array}\right.\ ] ] where @xmath103 and @xmath97 are given in and , respectively .
we are going to show that the answer to the @xmath0-dimensional matching problem is yes if and only if the constructed problem is feasible .
we first show that if the answer to the @xmath0-dimensional matching problem is yes , then problem is feasible .
in fact , if @xmath104 is a match for the @xmath0-dimensional matching problem , then a feasible power allocation of problem is given by @xmath105 2 , \quad \text{if}~n = l_z;\\[3pt ] 0 , \quad \text{if}~n\neq j_y~\text{or}~l_z .
\end{array}\right.\ ] ] this is because , since @xmath104 is a match for the @xmath0-dimensional matching problem , the above power allocation strategy is orthogonal to each other ( i.e. , @xmath106 ) .
furthermore , we have for user @xmath107 @xmath108 & \overset{(a)}{=}&\log_2\left(1+{\dfrac}{\alpha_{k_x , k_x}^{j_y}p_{k_x}^{j_y}}{\eta_{k_x}^{j_y}}\right)+\log_2\left(1+{\dfrac}{\alpha_{k_x , k_x}^{l_z}p_{k_x}^{l_z}}{\eta_{k_x}^{l_z}}\right)\\[20pt ] & \overset{(b)}{=}&\log_2\left(1+{\dfrac}{1 * 3}{1}\right)+\log_2\left(1+{\dfrac}{1 * 2}{2}\right)\\[10pt ] & = & 3,\end{array}\ ] ] where @xmath109 is due to and @xmath110 is due to , .
so is a feasible power allocation of problem . on the other hand , we show that if all the constraints in are satisfied , the answer to the @xmath0-dimensional matching problem must be yes .
notice that for any user @xmath111 , if the transmission rate @xmath112 , it must transmit on at least two subcarriers , for otherwise if it transmits only on one subcarrier , its maximum transmission rate is @xmath113 since there are @xmath3 users and @xmath114 subcarriers and at most one user is allowed to transmit on each subcarrier ( by the ofdma constraint ) , the feasibility of asks that each user in the network must transmit on _ exactly _ two subcarries . by the construction of the parameters of the system
, one can verify that the corresponding direct - link channel gains of the user on the two subcarriers must be @xmath115 , for otherwise the transmission rate is at most @xmath116 furthermore , the fact that all users transmission rate requirements are satisfied implies that each user transmits on one subcarrier in @xmath117 with noise power @xmath115 and one in @xmath118 with noise power @xmath119 which makes the transmission rate equal to @xmath120 otherwise , * if one user transmits on two subcarries in @xmath121 the transmission rate is at most @xmath122 * if one user transmits on two subcarries in @xmath117 both with noise power @xmath115 , then at least one user will transmit on two subcarries in @xmath118 either with noise power @xmath123 or @xmath0 and the transmission rate of this user is at most @xmath124 therefore , problem is feasible if and only if that for all users @xmath125 there exist @xmath126 and @xmath127 such that @xmath128 and @xmath129 according to the construction of @xmath130 and , we see that @xmath131 is a match for the @xmath0-dimensional matching problem .
it is simple to check that the above transformation from the @xmath0-dimensional matching problem to the feasibility problem can be done in polynomial time .
since the @xmath0-dimensional matching problem is strongly np - complete , we conclude that checking the feasibility of problem is strongly np - hard .
therefore , the optimization problem is also strongly np - hard . to illustrate the above proof , we take the following @xmath0-dimensional matching problem as an example , @xmath132@xmath133 @xmath134 and @xmath135 it is simple to check that @xmath136 is a match for the above given instance of the @xmath0-dimensional matching problem .
based on this @xmath0-dimensional matching problem , we construct a @xmath137-user @xmath138-carrier system with @xmath139 and @xmath140 . according to
, we have @xmath141 and @xmath142 the proof of lemma [ thm - feasi ] suggests the following system parameters ( cf .
) : all power budgets per subcarrier are @xmath143@xmath144all noise powers are @xmath145 except @xmath146@xmath147 and all direct - link channel gains are @xmath148 except @xmath149 \alpha_{3_x,3_x}^{2_y}=\alpha_{3_x,3_x}^{2_z}=\alpha_{3_x,3_x}^{4_y}=\alpha_{3_x,3_x}^{3_z}=\alpha_{4_x,4_x}^{3_y}=\alpha_{4_x,4_x}^{1_z}=1 .
\end{array}\ ] ] in this example , based on the match , we can construct an ofdma solution to the corresponding feasibility check problem , i.e. , @xmath150 \left(p_{2_x}^{1_y},p_{2_x}^{2_y},p_{2_x}^{3_y},p_{2_x}^{4_y},p_{2_x}^{1_z},p_{2_x}^{2_z},p_{2_x}^{3_z},p_{2_x}^{4_z}\right)=(3,0,0,0,0,2,0,0),\\[5pt ] \left(p_{3_x}^{1_y},p_{3_x}^{2_y},p_{3_x}^{3_y},p_{3_x}^{4_y},p_{3_x}^{1_z},p_{3_x}^{2_z},p_{3_x}^{3_z},p_{3_x}^{4_z}\right)=(0,0,0,3,0,0,2,0),\\[5pt ] \left(p_{4_x}^{1_y},p_{4_x}^{2_y},p_{4_x}^{3_y},p_{4_x}^{4_y},p_{4_x}^{1_z},p_{4_x}^{2_z},p_{4_x}^{3_z},p_{4_x}^{4_z}\right)=(0,0,3,0,2,0,0,0 ) .
\end{array}\ ] ] on the other hand , to look for a feasible solution of problem , we have to make each user transmit on two subcarriers ( one with noise power @xmath115 and the other with noise power @xmath123 ) and the direct - link channel gains of the user on the corresponding two subcarriers be @xmath151 notice that @xmath152 @xmath153 and @xmath154 we can ask user @xmath155 to transmit on subcarriers @xmath156 and @xmath157 user @xmath158 to transmit on subcarriers @xmath159 and @xmath160 user @xmath161 to transmit on subcarriers @xmath162 and @xmath163 and user @xmath164 to transmit on subcarriers @xmath165 and @xmath166 , respectively .
consequently , @xmath167 is a match for the given instance of the @xmath0-dimensional matching problem .
lemma [ thm - feasi ] shows that checking the feasibility of problem is strongly np - hard when @xmath168 based on this basic result , we can further prove that it is strongly np - hard to check the feasibility of problem when @xmath169 provided that @xmath170 is a strictly greater than one constant .
we summarize this result as theorem [ thm - feasi2 ] , and relegate its proof to appendix [ app - thm - feasi2 ] .
[ thm - feasi2 ] given any constant @xmath171 checking the feasibility of problem is strongly np - hard when @xmath169 .
thus , problem itself is also strongly np - hard .
remark 1 : problem remains strongly np - hard if the per - subcarrier power budget constraints @xmath172 there are replaced by the total power constraints @xmath173 or @xmath174 by setting @xmath175 or @xmath176 and using the same argument as in the proof of lemma [ thm - feasi ] and theorem [ thm - feasi2 ] , the strong np - hardness of the corresponding feasibility problems can be shown .
in fact , all strong np - hardness results of problems and in this paper also hold true for problems with either of the above two total power constraints .
remark 2 : another extension of problem is the so - called joint subcarrier and bit allocation problem @xcite . the goal of the joint subcarrier and bit allocation problem is to allocate subcarriers to users and at the same time allocate transmission bits to each user - subcarrier pair such that the total transmission power is minimized and the ofdma constraints and all users transmission requirements are satisfied .
mathematically , the problem can be formulated as @xmath177\text{s.t .
} & \displaystyle \sum_{n\in{{\cal n } } } r_k^n\geq \gamma_k,~k\in{{\cal k}},\\[15pt ] & \displaystyle p_k^n={(2^{r_k^n}-1)\eta_k^n}/{\alpha_{k , k}^n},~k\in{{\cal k}},~n\in{{\cal n } } , \\[5pt ] & r_k^n\in\left\{r_1,r_2, ... ,r_m,0\right\},~k\in{{\cal k}},~n\in{{\cal n}},\\[5pt ] & p_k^np_j^n=0,~\forall~j\neq
k,{~k,\,j\in{{\cal k}},}~n\in{{\cal n}}. \end{array}\]]the second constraint in says that @xmath178 ( cf . )
is necessary for user @xmath11 transmitting on subcarrier @xmath12 to achieve transmission rate @xmath179 , and the third constraint in enforces @xmath179 to take values in the possible transmission rate set @xmath180 given a total power budget @xmath181 , the decision version of problem is to ask whether there exists a feasible power and bit allocation strategy such that the optimal value of is less than or equal to @xmath182 it was shown in @xcite that the decision version of problem is strongly np - hard when @xmath183 .
the proof is based on a polynomial time transformation from the scheduling problem@xcite . in fact , by setting @xmath184 and using the same argument as in lemma 3.1 and theorem 3.1 , we can also show the strong np - hardness of the decision version of problem . in this subsection , we study the computational complexity of the system utility maximization problem for the multi - user ofdma system , where the utility is one of the four system utility functions introduced in section [ sec - model ] .
theorem [ thm - uti2 ] establishes the strong np - hardness of the problem for the general case @xmath61 .
the proof of theorem [ thm - uti2 ] can be found in appendix [ app - thm - uti2 ] .
[ thm - uti2]given any constant @xmath171 the system utility maximization problem with @xmath185 or @xmath186 is strongly np - hard when @xmath187 remark 3 : the result in theorem [ thm - uti2 ] is different from the one in @xcite .
it was shown in @xcite that the sum - rate utility maximization problem ( i.e. , @xmath188 ) is np - hard when the number of users is equal to @xmath124 the proof is based on a polynomial time transformation from the equipartition problem @xcite , which is known to be np - complete but not strongly np - complete .
theorem [ thm - uti2 ] shows that the sum - rate utility maximization problem is _ strongly _ np - hard .
the other three utility functions ( @xmath189 or @xmath186 ) are not considered in @xcite . the reference @xcite proved the np - hardness of the three utility maximization problems ( @xmath189 or @xmath186 ) in the two - user case by establishing a polynomial time transformation from the equipartition problem .
however , the proof of @xcite is nonrigorous , since the given equipartition problem has a yes answer does not imply that the transmission rate of the two users ( at the solution of the corresponding utility maximization problem ) is equal to each other , and vice versa .
the complexity status of problem with the proportional fairness utility , the harmonic mean utility , and the min - rate utility remains unknown when @xmath190 is fixed . in this section
, we have shown that both the total power minimization problem and the system utility maximization problem are strongly np - hard .
the basic idea of the proof is establishing a polynomial time transformation from the @xmath0-dimensional matching problem to the decision version of problems and .
the complexity result suggests that there are not polynomial time algorithms which can solve problems and to global optimality ( unless p@xmath2np ) .
therefore , one should abandon efforts to find globally optimal subcarrier and power allocation strategy for problems and , and determining an approximately optimal subcarrier and power allocation strategy is more realistic in practice .
in this section , we identify some easy cases when problem or problem can be solved in polynomial time . before doing this
, we introduce a concept called _ strong _ polynomial time algorithm .
a problem is said to admit a strong polynomial time algorithm if there exists an algorithm satisfying the following two conditions : * the complexity of the algorithm ( when applied to solve the problem ) depends only on the dimension of the problem and is a polynomial function of the dimension ; * the algorithm solves the problem to global optimality ( not just @xmath1-global optimality ) .
we remark that so far we do not know whether there exists a strong polynomial time algorithm to solve the general linear programming . when the interior - point algorithm is applied to solve the linear programming , it is only guaranteed to return an @xmath1-optimal solution in polynomial time and the complexity of the interior - point algorithm depends on the factor @xmath191@xcite .
the best complexity results of solving the linear programming is still related to the condition number of the constraint matrix @xcite . in the following subsections , we identify four ( strong ) polynomial time solvable subclasses of problems and .
more specifically , we first show that both problem and problem are ( strongly ) polynomial time solvable when there is only one user in the system ; see subsection [ sub - k=1 ] and subsection [ subsec - k=1 ] , respectively .
the ( extended ) `` water - filling '' technique plays a fundamental role in proving the polynomial time complexity .
then , we show in subsection [ subsecn = k ] that problem is strongly polynomial time solvable when the number of subcarriers is equal to the number of users . in this case
, we can reformulate problem as an assignment problem for a complete bipartite graph , which can be solved in strong polynomial time .
finally , we show the polynomial time complexity of problem with sum - rate utility without the total power constraint by transforming it into the polynomial time solvable hitchcock problem in subsection [ subsec - nopower ] .
when there is only one user ( i.e. , @xmath192 ) , problem becomes @xmath193\text{s.t .
} & \displaystyle\sum_{n\in{{\cal n}}}\log_2\left(1+{\dfrac}{\alpha^np^{n}}{\eta^n}\right)\geq \gamma , \\[15pt ] & p^n\geq p^n\geq 0,~n\in{{\cal n}}. \end{array}\ ] ] we claim that solving problem is equivalent to finding a minimal @xmath194 such that the optimal value of problem @xmath195\text{s.t . } & \displaystyle\sum_{n\in{{\cal n}}}p^n\leq p , \\[15pt ] & p^n\geq p^n\geq 0,~n\in{{\cal n}}\end{array}\ ] ] is equal to @xmath196 this is an important observation towards obtaining the closed - form solution of problem .
in fact , if the optimal value of problem is @xmath197 , then the optimal value of problem is @xmath198 and vice versa . in more details , for any fixed @xmath181 ,
problem is strictly convex with respect to @xmath199 and hence has a unique solution .
therefore , the set defined by @xmath200 & \displaystyle\sum_{n\in{{\cal n}}}p^n\leq { p^ * } , \\[15pt ] & p^n\geq p^n\geq 0,~n\in{{\cal n}}\end{array}\right.\ ] ] contains only one point , which must be the solution of problem .
hence , problems and share the same solution .
the solution to problem or problem is given by the following extended water - filling solution @xmath201where _
@xmath202 is chosen such that the objective value of problem is equal to @xmath203 _ and @xmath204 after obtaining the optimal @xmath202 , the optimal value @xmath197 of problem is given by @xmath205 for completeness , we show in appendix [ app-2 ] that indeed is the solution to problem and @xmath202 is actually the lagrangian multiplier corresponding to the constraint @xmath206 .
we point out that the water - filling solution extends the conventional water - filling solution @xmath207 in @xcite in the following two respects .
first , the conventional water - filling solution solves the power control problem without the power budget constraints per subcarrier , while the power control problem not only involves the total power constraint but also involves the power budget constraints per subcarrier .
second , the parameter @xmath202 in the conventional water - filling solution is chosen such that @xmath208 while the parameter @xmath202 in is chosen such that @xmath209 the only left problem now is to find @xmath210 in such that the objective value of problem is equal to @xmath196 a natural way to find the desired @xmath211 is to perform a binary search on @xmath212 since the objective function of is an increasing function with respect to @xmath213 as is known , the efficiency of the binary search depends on the initial search interval of @xmath213 to derive a good lower and upper bound , we first order the sequence @xmath214 and without loss of generality , suppose that @xmath215 notice that it takes @xmath216 operations to order @xmath217 then we calculate the objective values of problem at @xmath218 denoting them by @xmath219 .
it follows from the monotonicity of @xmath220 that @xmath221 if there exists an index @xmath222 such that @xmath223 then @xmath224 otherwise , we have @xmath225 for some @xmath222 and hence @xmath226 in this case , we can start the binary search from @xmath227 and it takes at most @xmath228 iterations to obtain an @xmath1-optimal @xmath213 assume that @xmath229 where @xmath230 is a sufficiently large constant .
then , we can conclude that problem is polynomial time solvable and the worst case complexity of solving it is @xmath231 remark 4 : the complexity status of the total power minimization problem remains unknown when @xmath190 is fixed .
if the system has only one user , all the four system utility functions coincide and problem becomes @xmath232\text{s.t .
} & \displaystyle \sum_{n\in{{\cal n}}}p^n\leq p , \\[15pt ] & p^n\geq p^n\geq 0,~n\in{{\cal n}}. \end{array}\ ] ] the main difference between problem and problem lies in that , @xmath194 is a given constant in problem , but is an unknown parameter in problem .
this feature of @xmath194 in problem enables us to design a strong polynomial time algorithm for the problem . without loss of generality , we assume that the parameters in problem satisfy @xmath233 otherwise , the solution to problem is @xmath234 for all @xmath235 notice that the solution to problem is given by the extended water - filling solution , where @xmath210 is chosen such that @xmath236 the desired @xmath237 here can be found by solving a univariate linear equation .
in fact , in a similar fashion as in subsection [ sub - k=1 ] , we can order @xmath238 in and assume that holds .
then we calculate the total transmission power @xmath239 at @xmath240 , denoting them by @xmath241 .
it follows from the monotonicity of @xmath220 that @xmath242 if there is an index @xmath243 such that @xmath244 , then @xmath245 .
otherwise , we have that @xmath246 for some @xmath247 . then for each @xmath248 we have @xmath249 = & \left\{\begin{array}{ll } p^n , \quad \text{if}~{\dfrac}{\eta^n}{\alpha^n}+p^n\leq b^{n^*};\\[10pt ] { \tau}-{\dfrac}{\eta^n}{\alpha^n } , \quad \text{if}~{\dfrac}{\eta^n}{\alpha^n } < b^{n^*}<{\dfrac}{\eta^n}{\alpha^n}+p^n;\\[10pt ] 0 , \quad \text{if}~{\dfrac}{\eta^n}{\alpha^n}\geq b^{n^*}. \end{array}\right . \end{aligned}\ ] ] therefore , the problem of finding the desired @xmath250 reduces to solve a univariate linear equation in terms of @xmath212 and the desired @xmath250 is obtained in a closed form . from the above discussion , we see that the complexity of finding the desired @xmath250 is @xmath251 remark 5 : the reference @xcite has addressed problem , but without power budget constraints per subcarrier .
the reference @xcite has also shown that the sum - rate maximization problem is np - hard when @xmath252 when the number of users is equal to the number of subcarriers ( i.e. , @xmath253 ) , we transform problem into an assignment problem for a complete bipartite graph with @xmath91 nodes in polynomial time .
specifically , we construct the bipartite graph @xmath254 with * the node set @xmath255 where @xmath6 and @xmath256 correspond to the set of users and the set of subcarriers , respectively ; * the edge set @xmath257 with the weight @xmath258 of edge @xmath259 @xmath260 \displaystyle\sum_{k\in{{\cal k}}}\sum_{n\in{{\cal n}}}p_k^n , \quad \text{otherwise}. \end{array}\right.\ ] ] therefore , problem can be equivalently reformulated as @xmath261 \text{s.t . }
& \displaystyle \sum_{k\in{{\cal k } } } x_{k}^n=1,~n\in{{\cal n } } , \\[12pt ] & \displaystyle \sum_{n\in{{\cal n } } } x_{k}^n=1,~k\in{{\cal k}},\\[12pt ] & x_{k}^n\in\left\{0,1\right\},~k\in{{\cal k}},~n\in{{\cal n}}. \end{array}\ ] ] in the above problem , the binary variable @xmath262 is equal to @xmath115 if user @xmath11 transmits power on subcarrier @xmath12 and @xmath263 otherwise .
the first constraint @xmath264 stands for the ofdma constraint , which requires that at most one user is allowed to transmit power on each subcarrier .
the second constraint @xmath265 requires that each user must transmit on one subcarrier to satisfy its specified transmission rate requirement . from (
* ? ? ? * theorem 11.1 ) , we know that the hungarian method solves problem to global optimality in @xmath266 operations .
the hungarian method is in essence a primal - dual simplex algorithm for solving the linear program @xmath267 \text{s.t . }
& \displaystyle \sum_{k\in{{\cal k } } } x_{k}^n=1,~n\in{{\cal
n } } , \\[12pt ] & \displaystyle \sum_{n\in{{\cal n } } } x_{k}^n=1,~k\in{{\cal k}},\\[12pt ] & 0\leq x_{k}^n\leq 1,~k\in{{\cal k}},~n\in{{\cal n } } , \end{array}\]]which is a relaxation of problem by replacing @xmath268 with @xmath269 in general , the primal - dual simplex method is not a polynomial time algorithm .
when it is used to solve problem , however , it can return the _ global optimal integer solution _ of problem in @xmath266 operations @xcite .
we remark that if the optimal value @xmath270 of problem satisfies @xmath271 the optimal solution to is @xmath272 otherwise , we will have that @xmath273which indicates that problem is infeasible .
remark 6 : in fact , problem is also polynomial time solvable when @xmath274 where @xmath275 is a given constant integer .
specifically , we can first partition @xmath276 subcarriers into @xmath3 nonempty subsets @xmath277 denote the number of ways to partition @xmath276 subcarriers into @xmath3 nonempty subsets by @xmath278 we show in appendix [ app - s ] that @xmath279 is upper bounded by @xmath280 after the partition of the subcarriers , the problem reduces to the case @xmath281 the only difference here is that the parameter @xmath282 is set to be @xmath283\text{s.t . } & \displaystyle \sum_{l\in{{\cal n}}_n}\log_2\left(1+{\dfrac}{\alpha_{k , k}^lp_k^{l}}{\eta_k^l}\right)\geq \gamma_{k } , \\[15pt ] & p_k^l\geq p_k^l\geq
0,~l\in { \cal n}_n\\ \end{array}\ ] ] if the above problem is feasible ; otherwise @xmath284 we know from subsection [ sub - k=1 ] that problem is polynomial time solvable under a mild assumption .
actually , if problem is feasible and @xmath285 only contains a singleton , we have @xmath286 which is the same as the one in .
then , for a given partition @xmath287 we can use the hungarian method to solve problem in @xmath266 operations , and assume the optimal value to be @xmath288 therefore , the optimal value of the original problem is @xmath289 if it is strictly less than @xmath290 otherwise the original problem is infeasible .
it follows from the above analysis that problem with @xmath291 can be solved in @xmath292 operations .
if we drop the total power constraint @xmath293 then problem with @xmath188 becomes @xmath294 & p_k^np_j^n=0,~\forall~j\neq k,{~k,\,j\in{{\cal k}},}~n\in{{\cal n}}. \end{array}\ ] ] problem is always feasible as @xmath36 ( for all @xmath11 and @xmath12 ) is a certificate for the feasibility .
further , since problem does not involve the total power constraint for each user , its optimal solution @xmath295 is either @xmath263 or @xmath296 to solve problem in polynomial time , we consider transforming it into the polynomial time solvable hitchcock problem @xcite ( also known as transportation problem ) . * hitchcock problem*. suppose there are @xmath3 sources of some commodity , each with a supply of @xmath297 units , and @xmath4 terminals , each of which has a demand of @xmath298 units .
suppose that the unit cost of transporting the commodity from source @xmath11 to terminal @xmath12 is @xmath299 the problem is how to satisfy the demands at a minimal cost ?
in fact , by setting @xmath300 @xmath301 @xmath302@xmath303 where @xmath304 we can see that problem is equivalent to the following hitchcock problem @xmath305 \text{s.t . } & \sum_{k=1}^k x_k^n=1,~n=1,2, ... ,n,\\[5pt ] & \sum_{k=1}^k x_k^{n+1}=(k-1)n,\\[5pt ] & \sum_{n=1}^n x_k^n = n,~k=1,2, ... ,k,\\[5pt ] & x_k^n=\left\{0,1\right\},~k=1,2, ...
,k,~n=1,2, ... ,n .
\end{array}\ ] ] in the above problem , the binary variable @xmath306 if user @xmath11 transmits full power @xmath34 on subcarrier @xmath12 and @xmath307 if user @xmath11 does not transmit any power on subcarrier @xmath12 .
the constraint @xmath308 and the constraint @xmath309 implies that at most one user is allowed to transmit power on each subcarrier @xmath12 .
however , one user is allowed to transmit on multiple subcarriers .
the variables @xmath310 are auxiliary dummy variables , since @xmath311 for all @xmath312 moreover , we know from ( * ? ? ?
* theorem 13.3 and its corollary ) that problem is equivalent to the linear program @xmath313 \text{s.t . }
& \sum_{k=1}^k x_k^n=1,~n=1,2, ... ,n,\\[5pt ] & \sum_{k=1}^k x_k^{n+1}=(k-1)n,\\[5pt ] & \sum_{n=1}^n x_k^n = n,~k=1,2, ... ,k,\\[5pt ] & 0\leq x_k^n\leq 1,~k=1,2, ... ,k,~n=1,2, ... ,n .
\end{array}\]]the equivalence between problems and is because the linear equality constraint in satisfies the so - called _ totally unimodular _
property@xcite , and thus all the vertices of the feasible set of problem are integer . since the linear program is polynomial time
solvable , the sum - rate maximization problem is polynomial time solvable .
further , suppose @xmath314 is the optimal solution to problem , then the optimal solution @xmath315 to problem is @xmath316 it is worthwhile remarking that the so - called @xmath317-algorithm ( * ? ? ?
* section 7.4 ) can efficiently solve problem , although it is not a polynomial time algorithm .
remark 7 : in a similar fashion , we can show that the _ weighted _ sum - rate maximization problem @xmath318 \text{s.t . } & p_k^n\geq p_k^n\geq 0,~k\in{{\cal k}},~n\in{{\cal n}},\\[5pt ] & p_k^np_j^n=0,~\forall~j\neq k,{~k,\,j\in{{\cal
k}},}~n\in{{\cal n}}\end{array}\]]is also polynomial time solvable , where @xmath319 are non - negative weights . again , we can transform problem into the hitchcock problem in polynomial time .
the corresponding parameters of the hitchcock problem are the same as before , except @xmath320 for @xmath321 and @xmath322 we summarize the main results in this section as the following theorem .
the following statements are true .
* problem is strongly polynomial time solvable when the number of users is equal to the number of subcarriers .
* problem is polynomial time solvable when there is only one user in the system .
* problem is strongly polynomial time solvable when there is only one user in the system . *
the ( weighted ) sum - rate maximization problem without the total power constraint is polynomial time solvable . in this section ,
we have identified four subclasses of problems and which are ( strongly ) polynomial time solvable . by doing so
, we successfully pick a subset of computationally tractable problems within the general class of strongly np - hard joint subcarrier and power allocation problems .
dynamic allocation of subcarrier and power resources in accordance with channel and traffic load changes can significantly improve the network throughput and spectral efficiency of the multi - user multi - carrier communication system where a number of users share some common discrete subcarriers .
a major challenge associated with joint subcarrier and power allocation is to find , for a given channel state , the globally optimal subcarrier and power allocation strategy to minimize the total transmission power or maximize the system utility function .
this paper mainly studies the computational challenges of the joint subcarrier and power allocation problem for the multi - user ofdma system .
we have shown that the general joint subcarrier and power allocation problem for the multi - user ofdma system is strongly np - hard .
the complexity result suggests that we should abandon efforts to find globally optimal subcarrier and power allocation strategy for the general multi - user ofdma system unless for some special cases ( i.e. , the case when there is only one user in the system , or the case when the number of users is equal to the number of subcarriers ) .
the problem is shown to be ( strongly ) polynomial time solvable in these special cases . in a companion paper
, we shall design efficient algorithms to solve the joint subcarrier and power allocation problem .
the first author wishes to thank professor zhi - quan ( tom ) luo of university of minnesota for inviting him to visit xidian university , where professor luo organized a summer seminar on optimization and its applications to signal processing and communications from may 21 to august 19 , 2012 .
this work was started from then .
the first author also would like to thank dr .
peng liu at xidian university and dr .
qiang li at chinese university of hong kong for many useful discussions .
before going into very details , let us first give a high level preview of the proof .
the basic idea of proving the strong np - hardness of problem with @xmath61 is to reduce it to the one with @xmath168 more specifically , we first partition all users into two types ( type - i and type - ii ) and also all subcarriers into two types ( type - i and type - ii ) , where the number of type - i subcarriers is required to be twice as large as the number of type - i users .
then , we construct `` good '' channel parameters between type - i ( type - ii ) users and type - i ( type - ii ) subcarriers while `` bad '' channel parameters between type - i ( type - ii ) users and type - ii ( type - i ) subcarriers such that the only way for all users to satisfy their transmission rate requirements is that type - i ( type - ii ) users will only transmit power on type - i ( type - ii ) subcarriers .
in addition , in our construction , all type - ii users and type - ii subcarriers are dummy ones , since all type - ii users transmission rate requirements can easily be satisfied by transmitting full power on type - ii subcarriers . in this way , the problem of checking the feasibility of problem with @xmath61 reduces to the one of checking whether all type - i users transmission rate targets can be met or not , where the number of type - i subcarriers is twice larger than the number of type - i users ( @xmath323 ) as required in the partition .
below is the detailed proof of theorem [ thm - feasi2 ] . by lemma [ thm - feasi ]
, we only need to consider the following two cases : ( i ) @xmath324 and ( ii ) @xmath325 . in case ( i ) , we partition @xmath3 users into @xmath326 type - i users and @xmath327 type - ii users , and @xmath328 subcarriers into @xmath329 type - i subcarriers and @xmath327 type - ii subcarriers .
we construct the channel parameters of type - i users on type - i subcarriers in the same way as in the proof of lemma [ thm - feasi ] where @xmath323 .
furthermore , noise powers , direct - link channel gains , and power budgets of all type - ii users on type - ii subcarriers are set to @xmath330 @xmath331 and @xmath332 respectively ; these parameters of all type - i users on type - ii subcarriers and all type - ii users on type - i subcarriers are set to @xmath333 and @xmath331 respectively .
all type - ii users desired transmission rate targets are set to be @xmath334 see fig .
[ plot ] for the corresponding ofdma system .
our construction is such that the channel condition of the type - i ( type - ii ) users on the type - i ( type - ii ) subcarriers is reasonably better than the one of the type - i ( type - ii ) users on the type - ii ( type - i ) subcarriers .
one can check that the only possible way for all @xmath3 users to meet their transmission rate targets is that , each type - ii user transmits full power on ( any ) one type - ii subcarrier ( actually type - ii users and subcarriers are dummy users and subcarriers ) and all type - i users appropriately transmit power on type - i subcarriers . by lemma [ thm - feasi ] , however , checking whether all type - i users transmission rate requirements can be satisfied is strongly np - hard . where @xmath323 .
noise powers , direct - link channel gains , power budgets of type - ii user(s ) on type - ii subcarriers are set to @xmath330 @xmath331 and @xmath332 respectively ; these parameters of type - i users on type - ii subcarriers and type - ii user(s ) on type - i subcarriers are set to @xmath333 and @xmath331 respectively . for case ( i ) , the transmission rate targets of type - ii users are all set to @xmath334 for case ( ii ) , the transmission rate target of the single type - ii user is set to @xmath335,width=366 ] in case ( ii ) , we partition @xmath3 users into @xmath336 type - i users and @xmath115 type - ii user , and @xmath328 subcarriers into @xmath337 type - i subcarriers and @xmath338 type - ii subcarriers .
we construct the channel parameters of type - i users on type - i subcarriers in the same way as in the proof of lemma [ thm - feasi ] where @xmath168 moreover , noise powers , direct - link channel gains , and power budgets of the single type - ii user on type - ii subcarriers are set to @xmath330 @xmath331 and @xmath332 respectively ; these parameters of all type - i users on type - ii subcarriers and the single type - ii user on type - i subcarriers are set to @xmath333 and @xmath331 respectively .
the transmission rate of the single type - ii user is required to be not less than @xmath335 see fig .
[ plot ] for the corresponding system . due to special construction of the system
, one can check that the only way for all @xmath3 users to meet their transmission rate targets is that , the single type - ii user transmits full power on all type - ii subcarriers and all type - i users appropriately transmit power on type - i subcarriers .
again , by lemma [ thm - feasi ] , checking whether all type - i users transmission rate requirements can be satisfied is strongly np - hard .
we first prove the strong np - hardness of problem for the special case @xmath323 and then prove their strong np - hardness for the general case @xmath339 * we first consider the case @xmath85 .
* for any instance of the @xmath0-dimensional matching problem , we construct the same system as in the proof of lemma [ thm - feasi ] and set @xmath340 * strong np - hardness of problem with @xmath341 * lemma [ thm - feasi ] directly implies that the following problem @xmath342 \text{s.t . } & \displaystyle \sum_{n\in{{\cal n}}}p_k^{n}\leq p_{k},~k\in{{\cal k } } , \\[12pt ] & p_k^n\geq p_k^n\geq 0,~k\in{{\cal k}},~n\in{{\cal n}},\\[5pt ] & p_k^np_j^n=0,~\forall~j\neq k,{~k,\,j\in{{\cal k}},}~n\in{{\cal n}}\end{array}\ ] ] is strongly np - hard , since the problem of checking whether its optimal value is greater than or equal to @xmath0 is strongly np - hard . * strong np - hardness of problem with @xmath343 * we prove that the sum - rate maximization problem @xmath344 \text{s.t . } & \displaystyle \sum_{n\in{{\cal n}}}p_k^{n}\leq p_{k},~k\in{{\cal k } } , \\[12pt ] & p_k^n\geq p_k^n\geq 0,~k\in{{\cal k}},~n\in{{\cal n}},\\[5pt ] & p_k^np_j^n=0,~\forall~j\neq k,{~k,\,j\in{{\cal k}},}~n\in{{\cal n}}\end{array}\ ] ] is also strongly np - hard . in particular , we show that checking the optimal value @xmath345 of problem is greater than or equal to @xmath0 is strongly np - hard , where @xmath346 is the transmission rate of user @xmath347 at the solution of problem . to this aim , consider the following relaxation of problem , @xmath348 \text{s.t . } & \displaystyle \sum_{k\in{{\cal k}}}\sum_{n\in{{\cal n}}}p_k^{n}\leq 5k , \\[13pt ] & p_k^n\geq 0,~k\in{{\cal k}},{~k,\,j\in{{\cal k}},}~n\in{{\cal n}},\\[5pt ] & p_k^np_j^n=0,~\forall~j\neq k,~n\in{{\cal n } } , \end{array}\ ] ] where @xmath349 due to the ofdma constraint , we see that problem is equivalent to @xmath350 \text{s.t . } & \displaystyle \sum_{n\in{{\cal n}}}p^{n}\leq 5k,\\[12pt ] & p^n\geq 0,~n\in{{\cal n}}. \end{array}\ ] ] noticing that problem is convex , we can obtain its optimal solution @xmath351 and its optimal value @xmath352 therefore , the optimal value @xmath345 of the original problem is less than or equal to @xmath332 and the equality holds if and only if @xmath353 the latter holds if and only if the answer to the @xmath0-dimensional matching problem is yes . therefore , the optimal value of problem is greater than or equal to @xmath0 if and only if the answer to the @xmath0-dimensional matching problem is yes .
* strong np - hardness of problem with @xmath354 and @xmath355 * for the cases @xmath356 and @xmath357 , notice that for all @xmath358 , @xmath359 , , @xmath360 @xmath361 and the equalities hold if and only if @xmath362 therefore , the optimal value of @xmath363 or @xmath364 is greater than or equal to @xmath0 if and only if the answer to the @xmath0-dimensional matching problem is yes .
this implies the strong np - hardness of all the four utility maximization problems with @xmath323 .
now we consider the general case @xmath365 we show the strong np - hardness of problem for the general case @xmath61 by constructing some dummy users and subcarriers as in the proof of theorem [ thm - feasi2 ] .
take the sum - rate maximization problem as an example .
it is simple to check that * in case @xmath366 the sum - rate utility function of the constructed system is greater than or equal to @xmath367 if and only if the given instance of the @xmath0-dimensional matching problem with size @xmath326 has a positive answer ; * in case @xmath171 the sum - rate utility function of the constructed system is greater than or equal to @xmath368 if and only if the given instance of the @xmath0-dimensional matching problem with size @xmath336 has a positive answer . hence , the sum - rate maximization problem in the case that @xmath61 is strongly np - hard .
similar results also hold true for the other three utility functions .
we omit the proof for brevity .
to show is the solution to problem , let us first write down the kkt condition of problem .
suppose @xmath369 are the lagrangian multipliers corresponding to the constraints @xmath370 @xmath371 and @xmath372 respectively , then the kkt condition of problem is given as follows @xmath373 & \displaystyle \sum_{n\in{{\cal n}}}p^n\leq p,~\lambda\geq 0,~\lambda\left(\sum_{n\in{{\cal n}}}p^n - p\right)=0,\\[15pt ] & p^n\geq p^n,~\xi^n\geq 0,~(p^n - p^n)\xi^n=0,~n\in{{\cal n}},\\[6pt ] & p^n\geq 0,~\nu^n\geq0,~p^n\nu^n=0,~n\in{{\cal n}}. \end{array}\right.\ ] ] since problem is convex for any fixed @xmath181 , it follows that the kkt condition is necessary and sufficient for @xmath374 to solve problem .
therefore , to show is the solution to problem , it suffices to show that there exist appropriate lagrangian multipliers such that @xmath375 satisfy the kkt system .
next , we construct appropriate @xmath376 such that @xmath374 in together with the constructed @xmath376 satisfy all the conditions in . specifically , we choose @xmath377 such that the objective value of problem is equal to @xmath203 and * if @xmath378 in , then @xmath379 we set @xmath380 and @xmath381 * if @xmath382 in , then @xmath383 we set @xmath384 * if @xmath234 in , then @xmath385 set @xmath386 and @xmath387 it is simple to check @xmath388 constructed in the above satisfies the kkt system . hence , is the solution to problem .
in combinatorics , a stirling number of the second kind @xcite , denoted by @xmath389 is the number of ways to partition a set of @xmath4 objects into @xmath3 nonempty subsets , where @xmath390 .
stirling numbers of the second kind obey the following recursive relation @xmath391 with initial conditions @xmath392 and @xmath393 for any @xmath394 to understand this formula , observe that a partition of @xmath395 objects into @xmath3 nonempty subsets either contains the @xmath396-th object as a singleton ( which corresponds to the term @xmath397 in ) or contains it with some other elements ( which corresponds to the term @xmath398 in ) .
next , we claim by induction that @xmath399 in fact , when @xmath400 we obtain @xmath401 this is because that dividing @xmath402 elements into @xmath3 sets means dividing it into one set of size @xmath123 and @xmath336 sets of size @xmath151 therefore , we only need to pick those two elements .
assume that @xmath403 we show that @xmath404 by invoking , we have@xmath405 = & s(k+c-1,k-1)+k{s(k+c-1,k)}\\[3pt ] \leq&s(k+c-1,k-1)+k{\left(k+c-1\right)}^{2(c-1)}\\[3pt ] \leq&s(k+c-1,k-1)+{\left(k+c\right)}^{2c-1}\\[3pt ] \leq&\sum_{k=1}^k \left(k+c\right)^{2c-1}\\[3pt ] \leq&(k+c)^{2c } , \end{array}\]]which shows that @xmath406 holds true .
h. dahrouj and w. yu , `` coordinated beamforming for the multicell multi - antenna wireless system , '' _ ieee trans .
wireless commun .
_ , vol . 9 , no .
. 17481759 , may 2010 . y .-
f . liu , y .- h .
dai , and z .- q .
luo , `` coordinated beamforming for miso interference channel : complexity analysis and efficient algorithms , '' _ ieee trans . signal process .
3 , pp . 11421157 , mar .
2011 .
f. bernardo , r. agust , j. cordero , and c. crespo , `` self - optimization of spectrum assignment and transmission power in ofdma femtocells , '' in _ proc .
advanced international conference on telecommunications _ , may , 2010 , pp . 404409 .
i. c. wong and b. l. evans , `` optimal downlink ofdma resource allocation with linear complexity to maximize ergodic rates , '' _ ieee trans .
wireless commun .
_ , vol . 7 , no .
962971 , mar .
2008 . s. hamouda , s. tabbane , and p. godlewski , `` improved reuse partitioning and power control for downlink multi - cell ofdma systems , '' in _ proc . international workshop on broadband wireless access for ubiquitous networking , _ sept .
2006 . l. m. c. hoo , b. halder , j. tellado , and j. m. cioffi , `` multiuser transmit optimization for multicarrier broadcast channels : asymptotic fdma capacity region and algorithms , '' _ ieee trans . commun .
52 , no . 6 , pp . 922930 , june 2004 .
c. y. wong , r. s. cheng , k. b. lataief , and r. d. murch , `` multiuser ofdm with adaptive subcarrier , bit , and power allocation , '' _ ieee j. sel .
areas in commun . , _ vol .
10 , pp . 17471758 ,
1999 .
s. v. hanly , l. l. h. andrew , and t. thanabalasingham , `` dynamic allocation of subcarriers and transmit powers in an ofdma cellular network , '' _ ieee trans .
theory , _ vol .
54455462 , dec . 2009 . c. mohanram and s. bhashyam , `` a sub - optimal joint subcarrier and power allocation algorithm for multiuser ofdm , '' _ ieee commun .
lett . , _ vol . 9 , no . 8 , pp . 685687 , aug .
k. e. baamrani , a. a. ouahman , v. p. g. jimnez , a. g. armada , and s. allaki , `` subcarrier and power allocation for the downlink of multiuser ofdm transmission , '' _ wireless personal communications , _ vol .
457465 , dec . 2006 .
l. tang , h. wang , q. chen , and g. m. liu , `` subcarrier and power allocation for ofdm - based cognitive radio networks , '' in _ proc .
ieee international conference on communications technology and applications , _ oct .
2009 , pp . 457461 .
m. s. alam , j. w. mark , and x. shen , `` relay selection and resource allocation for multi - user cooperative ofdma networks , '' _ ieee trans .
wireless commun . ,
5 , pp . 21932205 , may 2013 . m. dong , q. yang , f. fu , and k. s. kwak , `` joint power and subchannel allocation in relay aided multi - cell ofdma networks , '' in _ proc . international symposium on communications and information technologies , _ oct .
2011 , pp .
. g. huang , j. he , and q. zhang , `` research on adaptive subcarrier - bit - power allocation for ofdma , '' in _ proc .
international conference on wireless communications , networking and mobile computing , _ sept .
2009 , pp .
g. a. s. sidhu , f. gao , and a. nallanathan , `` a joint resource allocation scheme for multi - relay aided uplink multi - user ofdma system , '' in _ proc . international conference on wireless communications and signal processing , _ oct .
2010 , pp .
j. li , x. chen , c. botella , t. svensson , and t. eriksson , `` resource allocation for ofdma systems with multi - cell joint transmission , '' in _ proc .
ieee international workshop on signal processing advances in wireless communications , _ june 2012 , pp .
b. danobeitia , g. femenias , and f. riera - palou , `` an optimization framework for scheduling and resource allocation in multi - stream heterogeneous mimo - ofdma wireless networks , '' _ ifip , wireless days , _ nov .
2012 , pp . 1 - 3 .
f. wu , y. mao , x. huang , and s. leng , `` a joint resource allocation scheme for ofdma - based wireless networks with carrier aggregation , '' in _ ieee wireless communications and networking conference , _ apr .
2012 , pp . 12991304 .
m. belleschi , p. detti , and a. abrardo , `` complexity analysis and heuristic algorithms for radio resource allocation in ofdma networks , '' in _ proc .
international conference on telecommunications _ , may 2011 , pp .
. w. jiang , z. zhang , x. sha , and l. sun , `` low complexity hybrid power distribution combined with subcarrier allocation algorithm for ofdma , '' _
j. systems engineering and electronics , _ vol .
22 , no . 6 , pp .
879884 , dec .
r. o. afolabi , a. dadlani , and k. kim , `` multicast scheduling and resource allocation algorithms for ofdma - based systems : a survey , '' _ ieee communications surveys & tutorials , _ vol .
240254 , first quarter 2013 .
k. sumathi and m. l. valarmathi , `` resource allocation in multiuser ofdm systems a survey , '' in _ proc .
international conference on computing communication & networking technologies , _
july 2012 , pp .
liu , m. hong , and y .- h .
dai , `` max - nin fairness linear transceiver design problem for a multi - user simo interference channel is polynomial time solvable , '' _ ieee signal process .
2730 , jan . 2013 .
luo and s. zhang , `` duality gap estimation and polynomial time approximation for optimal spectrum management , '' _ ieee trans .
signal process .
_ , vol .
57 , no .
7 , pp . 26752689 , jul . 2009
. w. yu and r. lui , `` dual methods for nonconvex spectrum optimization of multicarrier systems , '' _ ieee trans .
54 , no . 7 , pp .
13101322 , july 2006 .
r. cendrillon , m. moonen , j. verliden , and t. bostoen , and w. yu , `` optimal multiuser spectrum management for digital subscriber lines , '' in _ proc .
ieee international conference on communications _ ,
june 2004 , vol .
1 , pp . 15 .
t. m. cover and j. a. thomas , _ elements of information theory , _
new york , u.s.a . :
john wiley & sons , inc .
a. ben - tal and a. nemirovski , _ lectures on modern convex optimization .
_ philadelphia , u.s.a .
: siam - mps series on optimization , siam publications , 2001 .
s. a. vavasis and y. ye , a primal - dual interior point method whose running time depends only on the constraint matrix , " _ math . prog .
74 , pp . 79120 , 1996 .
a. mohr and t. d. porter , `` applications of chromatic polynomials involving stirling numbers , '' _ journal of combinatorial mathematics and combinatorial computing , _ vol .
5764 , 2009 . | consider a multi - user orthogonal frequency division multiple access ( ofdma ) system where multiple users share multiple discrete subcarriers , but at most one user is allowed to transmit power on each subcarrier . to adapt fast traffic and channel fluctuations and
improve the spectrum efficiency , the system should have the ability to dynamically allocate subcarriers and power resources to users . assuming perfect channel knowledge , two formulations for the joint subcarrier and power allocation problem are considered in this paper : the first is to minimize the total transmission power subject to quality of service constraints and the ofdma constraint , and the second is to maximize some system utility function ( including the sum - rate utility , the proportional fairness utility , the harmonic mean utility , and the min - rate utility ) subject to the total transmission power constraint per user and the ofdma constraint . in spite of the existence of various heuristics approaches , little
is known about the computational complexity status of the above problem .
this paper aims at filling this theoretical gap , i.e. , characterizing the complexity of the joint subcarrier and power allocation problem for the multi - user ofdma system .
it is shown in this paper that both formulations of the joint subcarrier and power allocation problem are strongly np - hard . the proof is based on a polynomial time transformation from the so - called @xmath0-dimensional matching problem .
several subclasses of the problem which can be solved to global optimality or @xmath1-global optimality in polynomial time are also identified .
these complexity results suggest that there are not polynomial time algorithms which are able to solve the general joint subcarrier and power allocation problem to global optimality ( unless p@xmath2np ) , and determining an approximately optimal subcarrier and power allocation strategy is more realistic in practice
. computational complexity , power control , ofdma system , subcarrier allocation , system utility maximization . |
electromagnetic decays of vector mesons are a good probe of the hot and dense state created after a nucleus - nucleus collision .
indeed , the emitted lepton pairs , which carry the spectral information of the vector meson at the moment of the decay , undergo a negligible distortion as they travel through the dense medium @xcite .
this fact can in turn tell us about properties of the medium .
we consider interactions with pions ( @xmath1 ) and nucleons ( @xmath2 ) as scatterings of the @xmath0 meson that contribute to its self - energy . in this approach the forward scattering amplitude ( fsa )
is composed of a low energy and a high energy part .
the latter is described by regge parameterization whereas former encompasses two terms , namely a resonance term modeled by breit - wigner functions and background term . in the center of mass ( c.m . )
frame the low energy fsa is written as : @xmath3 here the sum ranges over all breit - wigner resonances that decay into the @xmath0 meson and the particle @xmath4 which is either a nucleon or a pion .
we now briefly explain the different variables involved .
the mass of the resonance @xmath5 is @xmath6 and its total width is @xmath7 .
@xmath8 is the usual mandelstam variable and the magnitude of the c.m .
momentum is given by : @xmath9 \left [ s - \left ( m_\phi - m_a \right)^2 \right ] } } { \sqrt{s } } \label{eq : qcm}\ ] ] where @xmath10 .
@xmath11 are statistical spin / isospin averaging factors : @xmath12 where @xmath13 ( with @xmath14 generic ) is the spin of particle @xmath14 and @xmath15 is the isospin of that particle .
the effective width @xmath16 takes the form @xcite : @xmath17 where @xmath18 is the total width of the resonance @xmath5 , @xmath19 is the branching ratio of the decay @xmath20 , @xmath21 \left [ m^{2}_{r } - \left ( m_\phi - m_a \right)^2 \right ] } } { m_{r } } $ ] , and @xmath22 is the minimal relative angular momentum between @xmath0 and @xmath4 .
the case of the pions is a bit more complicated for the @xmath23 case however , as adler s theorem should be fulfilled .
that is , the pion scattering amplitude on any hadronic target vanishes when @xmath24 in the limit of massless pions @xcite .
consequently , the effective width for @xmath23 now must be rewritten as : @xmath25 with @xmath26 .
finally , the high energy fsa is well described by a regge parametrization of the form : @xmath27 r_{\phi a}^{i } s^{\alpha_i}. \label{eq : f_high}\ ] ] where @xmath4 can be a nucleon or a pion as mentioned before .
the variables @xmath28 and @xmath29 are regge residues and intercepts which are determined experimentally @xcite .
the real and imaginary parts of the fsa are plotted in fig.[pic : re_im_f ] .
these plots are made relative to the rest frame of the heath bath . [ cols="^,^ " , ]
the results show that the mass shift overall is small , reaching 38 mev at most .
there is a significant broadening of the width of the @xmath0 meson however . at normal nucleon density and temperature of 150mev ,
the width more than 90 mev .
our results are important for understanding possible in - medium modifications of the @xmath0 meson strongly interacting environments .
studies to estimate phenomenological implications are in progress @xcite .
+ + 50 v. l. eletsky , m. belkacem , p. j. ellis , and j. i. kapusta , phys .
c * 64 * , 035202 ( 2001 ) .
a. t. martell and p. j. ellis , phys .
c * 69 * , 065206 ( 2004 ) .
kapusta , and c. gale , _ finite temperature field theory principles and applications _ ( cambridge university press , cambridge , uk , 2006 ) . w .-
et al . _ , j. phys .
g * 33 * , 1 ( 2006 ) .
g. vujanovic , c. gale , and j. ruppert , in preparation . | 1.0 cm 1.0 cm we calculate the spectral density of the @xmath0 meson in a hot bath of nucleons and pions using a general formalism relating self - energy to the forward scattering amplitude ( fsa ) @xcite .
in order to describe the low energy fsa , we use experimental data along with a background term . for the high energy fsa , a regge parameterization is employed .
we verify the resulting fsa using dispersion techniques .
we find that the position of the peak of the spectral density is slightly shifted from its vacuum position and that its width is considerably increased .
the width of the spectral density at a temperature of 150 mev and at normal nuclear density is more than 90 mev . |
nonlinear resonance is a problem with long history in quantum and classical mechanics , going back to laplace and poincare on the classical side and to the fermi resonance on the quantum side @xcite .
it occurs in a broad range of systems , from celestial bodies to ecological systems to molecules @xcite .
recently nonlinear resonance has attracted particular interest in the context of nano - and micromechanical vibrational systems @xcite as well as the microwave cavities used in quantum information .
these systems provide an unprecedented access not just to studying this complicated phenomenon , but also to controlling and using it . in conservative classical systems
, nonlinear resonance leads to the onset of nonlinear vibrations accompanied by energy oscillations between the resonating modes .
this is reminiscent of linear resonance between two harmonic oscillators .
however , on a finer scale , the picture is more complicated , extending to dynamical chaos . on the quantum side ,
nonlinear resonance is in some sense simpler in the absence of dissipation , as it leads to the standard level repulsion .
the quantum situation changes if the resonating modes are dissipative .
if the modes have very different decay rates , one of them can serve as a thermal reservoir for another @xcite . this effect has attracted much attention in cavity optomechanics @xcite and has been recently used to drive a slowly decaying microwave cavity mode to a coherent quantum state @xcite . however ,
this behavior can change dramatically when the mode nonlinearity is significant @xcite .
an important advantageous feature of mesoscopic oscillators is the possibility to tune the modes in and out of nonlinear resonance .
this can be done statically , by directly controlling the mode frequencies @xcite , and also dynamically , by taking advantage of the dependence of the mode eigenfrequency on the vibration amplitude . by driving a mode close to its eigenfrequency , and thus increasing its vibration amplitude
, one can bring its overtone in resonance with an overtone of another mode .
this second mode is then excited , too .
the backaction from this excitation can significantly change the dynamical response of the driven mode . in this paper
we develop a theory of weakly damped vibrational modes brought into nonlinear resonance .
a specific feature of the nonlinear resonance is that both the strength of the nonlinear coupling and the mode frequencies depend on the modal amplitudes .
therefore a fast decaying mode can be an efficient thermal reservoir for another mode in one range of the amplitudes , but not in the other .
one can then expect a strongly non - exponential decay of the slower - decaying mode . on the other hand ,
if in a certain amplitude range the decay is weaker than the coupling in the appropriate units , the dynamics corresponds to resonant inter - mode energy oscillations .
however , as the amplitudes decay , the scaled coupling can become weaker than the decay , eliminating the oscillations .
the damping of mesoscopic modes is usually weak , the decay rate is small compared to the mode frequencies .
such modes can be efficiently excited by a comparatively weak driving if it is resonant .
this is advantageous , as it allows keeping the leading terms responsible for the resonant behavior while eliminating irrelevant variables .
respectively , we will describe the dynamics of the coupled modes in the rotating wave approximation ( rwa ) .
our study aims at addressing two problems : the decay of modes that go through nonlinear resonance and the response of the modes to a resonant field .
we will concentrate on the limiting cases .
the general analysis can be only done numerically .
extensive numerical simulations have been done in parallel with this paper , along with the experimental studies simultaneously and independently by two groups @xcite . to be specific
, we will consider the modes with frequency ratio close to 1:3 .
in many nano- and micromechanical systems the coupling between such modes is comparatively strong ; for symmetry reasons , it is often stronger than the coupling between the modes close to 1:2 resonance .
the onset of 1:3 nonlinear resonance in a micromechanical system , which is similar to the one we use here , was inferred previously from indirect data , but the related nonlinear dynamics was not explored @xcite .
the work @xcite also refers to 1:3 resonance .
one of the specific limiting cases we will address are the case where one of the modes has a much higher decay rate than the other . in this case the dynamics can be described in the adiabatic approximation , in which the fast decaying mode follows the evolution of the slowly decaying mode adiabatically .
this leads to an anomalous nonlinear friction for the slowly decaying mode .
along with the nonlinear friction there arises a change of the effective vibration frequency of the slowly decaying mode , that displays a pronounced and strongly nonmonotonic dependence on the vibration amplitude .
this is the adiabatic analog of the nonlinear frequency anticrossing .
the other specific case is the case where the decay rates of the both modes are small compared to the appropriately scaled nonlinear mode coupling and possibly to the nonlinear shift of the mode frequencies compared to the small - amplitude eigenfrequency values . in this case
the dynamics in the absence of the drive and dissipation is fully integrable , in the rwa . in the rotating frame
, it maps onto vibrations of a pendulum .
" we will describe the effect of weak dissipation and a weak field , which in this case can lead to the onset of limit cycles in the rotating frame , with the amplitude determined by the competition between energy drain due to the dissipation and the energy gain from the field .
to explore the driving - tuned nonlinear resonance it suffices to study two coupled modes .
the modes can be described by the simplest model that displays the amplitude dependence of the vibration frequency , the duffing oscillator . in the absence of coupling to the environment that leads to dissipation and fluctuations
, the hamiltonian of the modes reads @xmath0 with @xmath1 here , @xmath2 and @xmath3 are the coordinates and momenta of the modes , @xmath4 are the mode eigenfrequencies , @xmath5 and @xmath6 are the parameters of the duffing nonlinearity , @xmath7 and @xmath8 are the amplitude and frequency of the driving field , and @xmath9 is the mode coupling parameter .
the model ( [ eq : hamiltonian ] ) captures the essential features of the resonance . in the considered regime , other nonlinear terms
renormalize the parameters .
we assume that @xmath8 is close to the mode-1 eigenfrequency , @xmath10 .
we also assume that @xmath11 is close to @xmath12 . however , the detuning @xmath13 , even though it is small compared to @xmath14 , is not necessarily small compared to the coupling strength , so that one needs the driving to tune modes in resonance .
the driving - induced tuning occurs because the frequency of the mode 1 depends on its amplitude , @xmath15 , where the bar means averaging over the vibration period .
if @xmath16 then , by increasing @xmath17 , one can make @xmath18 coincide with @xmath12 .
for concreteness we set @xmath19 ; respectively , of interest is the range where the frequency detuning of the field @xmath20 , so that the frequency of the forced vibrations of mode 1 could resonate with @xmath21 . to analyze the effect of resonant driving
, we can use the rotating wave approximation ( rwa ) and change from the coordinates and momenta of the modes to new scaled coordinates and momenta , @xmath22 with the phases @xmath23 and @xmath24 .
we use subscript @xmath25 to enumerate the dynamical variables of the modes 1 , 2 .
functions @xmath26 are slowly varying in time , they remain almost constant over the driving period @xmath27 .
the equations of motion for these functions are @xmath28 \nonumber\\ & -\frac{f}{2\sqrt{\omega_f}}q_1 + \frac{\gamma_{12}}{8\sqrt{3}\omega_f^2}{\rm re}\,[(q_1-ip_1)^3(q_2+ip_2)]\end{aligned}\ ] ] here , we have incorporated the terms proportional to @xmath29 and @xmath30 that describe the decay of the modes . in the phenomenological model of the mode dynamics they correspond to the friction forces @xmath31 that drive the modes ; incorporating them into eq .
( [ eq : rwa ] ) is justified even where the friction is delayed in the lab frame .
the frequency detunings are @xmath32 and we have disregarded corrections @xmath33 . a major feature of the driving - induced nonlinear resonance seen from eq .
( [ eq : rwa ] ) is that it can be described by autonomous equations of motion .
they have a hamiltonian part , with the hamiltonian @xmath34 , and a part that describes dissipation .
the coupling between the modes is described by the last term in @xmath34 .
the dynamics of coupled modes becomes particularly simple if mode 2 decays much faster than mode 1 , @xmath35 .
physically , mode 2 can serve as a thermal reservoir for mode 1 .
this reservoir " has two important features .
first , it has a finite bandwidth . in the considered 1:3 resonance ,
the reservoir is efficient if the vibration frequency of mode 1 is close to @xmath21 .
second , the coupling to the reservoir is nonlinear in the mode-1 coordinate
. therefore one may expect that the resulting dissipation rate will depend on the amplitude of mode 1 .
if the appropriately weighted mode coupling is small compared to @xmath30 and @xmath35 , mode 2 follows mode 1 adiabatically .
the restriction on the coupling is the analog of the familiar condition of a weak coupling to a thermal reservoir . because of that
, the vibration amplitude of mode 2 is comparatively small .
one can then disregard the internal duffing nonlinearity of mode 2 , i.e. , to set @xmath6 equal to zero in eq .
( [ eq : rwa ] ) . to gain insight into the ensuing nonlinear friction of mode 1
we will consider first the dynamics in the absence of driving , @xmath36 .
the problem here is how mode 1 decays from the initial value of its amplitude in the presence of resonant nonlinear coupling .
it is convenient to introduce a dimensionless complex amplitude of mode 1 , @xmath37e^{-i\phi_{\rm ad}(t)},\nonumber\\ & \phi_{\rm ad}(t ) = ( \gamma_2/3)\int^t dt ' |a_{\rm ad}(t')|^2 - \delta\omega_1 t.\end{aligned}\ ] ] in the adiabatic regime , one can solve the equation of motion for mode 2 by setting @xmath38 $ ] and disregarding @xmath39 compared to @xmath40 .
this gives a simple linear algebraic equation for @xmath41 .
using its solution , we obtain from eq .
( [ eq : rwa ] ) @xmath42 this expression applies for @xmath43 , a condition that imposes a limitation on the vibration amplitude and the coupling strength . for the period - averaged displacement @xmath17 of mode 1 it reads @xmath44 .
equation ( [ eq : adiabatic ] ) shows that the backaction from the fast - decaying mode 2 leads to a nonexponential decay of the vibration amplitude of mode 1 .
the decay rate is maximal where the amplitude - dependent frequency of mode 1 , @xmath45 , coincides with @xmath21 , in which case @xmath46 .
one can understand the mode - coupling induced relaxation in terms of the transitions into a thermal reservoir mimicked by mode 2 .
the energy relaxation rate of mode 1 , which is @xmath47 , has a term determined by the rate of such transitions .
this rate is given by the fermi golden rule .
the effective density of states of the reservoir at the relevant frequency @xmath18 is @xmath48 $ ] .
the transition rate is proportional to this density of states and is quadratic in the coupling @xmath49 .
the fermi - golden - rule calculation is in full agreement with eq .
( [ eq : adiabatic ] ) .
a nonexponential decay of the vibration amplitude is associated with a nonlinear friction force experienced by the vibrational system .
nonlinear friction is a topic of increasing interest in nano- and micro - mechanics , cf .
usually it is described by a term @xmath50 in the right - hand side of eq .
( [ eq : adiabatic ] ) .
the fact that nonlinear coupling to a fast decaying mode can lead to such nonlinear friction was indicated , but not elaborated in ref .
@xcite . in the present case
the nonlinear friction is anomalously large and has a different dependence on the mode amplitude .
if the duffing nonlinearity parameter @xmath5 is small , for not too large vibration amplitudes one can disregard the term @xmath51 in the denominator in eq .
( [ eq : adiabatic ] ) .
the decay of the the scaled squared vibration amplitude @xmath52^{1/2 } |a_{\rm ad}|^2 $ ] is then described by a simple expression @xmath53\}^{-1/2}\end{aligned}\ ] ] as expected , the decay is profoundly nonexponential in time .
it is faster than exponential , and becomes exponential only for large time , where @xmath54 . for larger initial amplitudes or stronger duffing nonlinearity the resonant nature of the nonlinear decay described by eq .
( [ eq : adiabatic ] ) becomes more pronounced .
it is illustrated in fig .
[ fig : adiabatic_amplitude ] for the scaled instantaneous decay rate @xmath55 .
the figure shows two important features of the instantaneous decay rate .
first , it displays a resonant peak where the instantaneous frequency of mode 1 @xmath56 equals to @xmath21 .
the height of the peak increases with the increasing coupling strength @xmath57 and with the increasing frequency detuning @xmath58 .
second , the instantaneous rate becomes constant not only for small vibration amplitudes , where @xmath59 , but also for @xmath60 , where @xmath61 . on the scaled vibration amplitude @xmath62 of mode 1 in the adiabatic regime of fast decaying mode 2 .
curves 1 and 2 refer to @xmath63 and @xmath64 and 0.4 , respectively .
curve 3 refers to @xmath65 and @xmath66,width=240 ] interestingly , for relatively large amplitudes , where @xmath67 , the detuning of the triple effective frequency of mode 1 , @xmath68 , largely exceeds the width @xmath30 of the resonant response of mode 2 .
this means that @xmath18 is on the tail of the lorentzian
density of states " of the effective thermal reservoir represented by mode 2 .
yet , the decay rate @xmath69 does not fall off to the bare " decay rate @xmath29 in the absence of the mode coupling .
this is a consequence of the strong nonlinearity of the coupling , with the coupling energy @xmath70 . along with dissipation , the backaction from a thermal reservoir generally leads , in quantum terms , to a change of the energy spectrum . in classical terms
, it corresponds to a change of the effective vibration frequency , @xmath71 .
this change in the adiabatic case is described by the imaginary part of @xmath72 in eq .
( [ eq : adiabatic ] ) , @xmath73^{-1}\}.\end{aligned}\ ] ] the shift @xmath74 is a strongly nonlinear function of the amplitude , much stronger than the duffing nonlinearity - induced shift , which is @xmath51 . in the interesting case @xmath75 , the shift @xmath74 changes sign with the varying vibration amplitude .
the sign change occurs at resonance , @xmath76 .
it is an analog of the level repulsion in quantum mechanics , which in this case looks like a turnaround of the vibration frequency .
it strongly affects the response of the mode 1 to the driving . somewhat unexpectedly , for @xmath75 the slope of the frequency @xmath77 as function of the amplitude can change twice with the increasing @xmath78 .
the first change , from the positive to the negative slope , can occur for @xmath79 provided the corresponding @xmath78 is small compared to @xmath80 , i.e. , for sufficiently strong initial detuning from the resonance ( this is the most interesting case ; the expression in the general case is elementary , but cumbersome ) .
then , with the increasing amplitude , the slope changes sign again before @xmath78 becomes larger than @xmath81 . in the presence of driving
, this unusual behavior leads to an unusual dependence of the amplitude @xmath62 on the parameters of the driving field . with the driving on ,
in the limit of fast decay of mode 2 the equation for @xmath82 reads @xmath83 where @xmath84 is the scaled squared force amplitude ; functions @xmath85 and @xmath86 are defined in eqs .
( [ eq : amplitudes ] ) and ( [ eq : adiabatic ] ) . because of the nonlinear friction and the nonmonotonic dependence of the frequency on amplitude , the dynamics described by eq .
( [ eq : adiabatic_drive ] ) is significantly more complicated than the dynamics of a duffing oscillator . in dimensionless time
@xmath87 it depends on five parameters .
it is important that all parameters , except for the coupling constant @xmath9 , can be independently determined in the experiment by studying the dynamics of the modes for comparatively small drive , where the modes are essentially uncoupled .
one of the first steps toward the understanding of the features of the dynamics related to the coupling is to study the stationary solution of eq .
( [ eq : adiabatic_drive ] ) .
it determines how the amplitude and phase of forced vibrations vary with the varying amplitude @xmath7 and frequency @xmath8 of the drive .
the amplitude of the forced vibrations is particularly important , as this is usually the first characteristic measured in the experiment .
equation ( [ eq : adiabatic_drive ] ) has a form of the polynomial equation for the squared vibration amplitude of the stationary state @xmath88 , @xmath89 the solution of this equation as function of the scaled field amplitude @xmath90 and the scaled frequency detuning @xmath91 is shown in fig .
[ fig:3d_amplitude ] ( a ) . as seen from fig .
[ fig:3d_amplitude ] , the amplitude @xmath92 has multiple branches and takes on different values for the same values of the drive parameters .
the pattern becomes increasingly complicated as the system is brought into the nonlinear resonance by the increasing drive .
this behavior is closely related to the nonmonotonous dependence of the `` vibration frequency '' @xmath77 on the vibration amplitude discussed above and also to the strongly nonlinear friction . and the scaled driving amplitude , @xmath93 .
other parameters are @xmath94 , and @xmath95 .
function @xmath62 for particular parameter values is illustrated in ( b ) and ( c ) .
( b ) frequency dependence of the amplitude for @xmath96 .
black curves correspond to stable stationary states of forced vibrations , red curves show unstable stationary states .
( c ) the vibration amplitude as a function of the drive amplitude for a given frequency detuning @xmath97 const .
curves 1 , 2 , and 3 refer to @xmath98 and 40 . as in ( b )
, black curves correspond to stable stationary states of forced vibrations , red curves show unstable stationary states .
the results for extremely large @xmath90 illustrate the bifurcations in the system ; they are of interest for the modes with extremely small decay rates @xmath99,title="fig:",width=268 ] + and the scaled driving amplitude , @xmath93 .
other parameters are @xmath94 , and @xmath95 .
function @xmath62 for particular parameter values is illustrated in ( b ) and ( c ) .
( b ) frequency dependence of the amplitude for @xmath96 .
black curves correspond to stable stationary states of forced vibrations , red curves show unstable stationary states .
( c ) the vibration amplitude as a function of the drive amplitude for a given frequency detuning @xmath97 const .
curves 1 , 2 , and 3 refer to @xmath98 and 40 . as in ( b ) , black curves correspond to stable stationary states of forced vibrations , red curves show unstable stationary states .
the results for extremely large @xmath90 illustrate the bifurcations in the system ; they are of interest for the modes with extremely small decay rates @xmath99,title="fig:",width=153 ] and the scaled driving amplitude , @xmath93 .
other parameters are @xmath94 , and @xmath95 .
function @xmath62 for particular parameter values is illustrated in ( b ) and ( c ) .
( b ) frequency dependence of the amplitude for @xmath96 .
black curves correspond to stable stationary states of forced vibrations , red curves show unstable stationary states .
( c ) the vibration amplitude as a function of the drive amplitude for a given frequency detuning @xmath97 const .
curves 1 , 2 , and 3 refer to @xmath98 and 40 . as in ( b )
, black curves correspond to stable stationary states of forced vibrations , red curves show unstable stationary states .
the results for extremely large @xmath90 illustrate the bifurcations in the system ; they are of interest for the modes with extremely small decay rates @xmath99,title="fig:",width=153 ] a part of the branches of @xmath92 correspond to stable states , whereas the other are unstable stationary states .
equation ( [ eq : adiabatic_drive ] ) describes the dynamics of a single complex variable or equivalently , two real variables .
the stationary states of the system are either stable states or saddle points . indeed , if we linearize eq .
( [ eq : adiabatic_drive ] ) about a stationary state , we will see that the sum of the two eigenvalues of the corresponding characteristic equation is negative , which indicates that at least one of the eigenvalues that characterize the linearized motion has a negative real part .
the system can switch between different stable states with varying parameters of the drive ( or because of noise ) , but in the adiabatic limit there are no stationary states where the amplitude @xmath82 would oscillate .
the stable and unstable branches of @xmath92 are shown in fig .
[ fig:3d_amplitude ] ( b ) and ( c ) . as seen from fig .
[ fig:3d_amplitude ] ( b ) , as the driving frequency @xmath8 is varied , the amplitude can fold over , and there emerge regions of coexistence of 2 or 3 stable states of forced vibrations .
these regions are separated by a gap where eq .
( [ eq : adiabatic_drive ] ) has only one stationary solution .
the gap is a consequence of the repulsion " of the vibration frequencies away from the resonance where @xmath100 .
it leads to an interesting hysteretic behavior .
if we start from negative @xmath91 and increase @xmath8 , we move from the small amplitude branch until it ends and we jump to the high - amplitude branch .
from there , if we keep increasing @xmath8 , we jump to the small - amplitude branch which goes to large positive @xmath101 .
the large amplitude branch for positive @xmath101 can not be accessed by varying @xmath8 at all .
a complicated character of the hysteresis with varying field strength for constant @xmath8 is seen from fig .
[ fig:3d_amplitude ] ( c ) . on branch 1 , which refers to the driving frequency in the region of the left
beak " in fig .
[ fig:3d_amplitude ] ( b ) , there are 3 coexisting stable states ; the system switches with varying @xmath7 between the large / small amplitude branch and the intermediate - amplitude branch .
branch 2 refers to the gap in fig .
[ fig:3d_amplitude ] ( b ) , where the system has only one branch .
branch 3 refers to the region of @xmath8 where there emerges the right beak " in fig .
[ fig:3d_amplitude ] ( b ) and the system has two branches of coexisting stable states .
the merging of the unstable and low - amplitude branches occurs for very large @xmath90 and is not shown .
the dynamics of resonating modes is interesting also in the opposite limit where the decay of the modes can be disregarded all together . as seen from eq .
( [ eq : rwa ] ) , such dynamics is described by a time - independent hamiltonian @xmath34 .
it is convenient to re - write the equations of motion for scaled variables @xmath102^{1/2}(q_j - ip_j ) \quad(j=1,2).\end{aligned}\ ] ] in dimensionless time @xmath103 in the neglect of the decay eqs .
( [ eq : rwa ] ) read @xmath104 with @xmath105 , @xmath106 and @xmath107 in the absence of the driving field @xmath108 , eqs .
( [ eq : energy_reduced ] ) provide a simple example of the standard nonlinear resonance .
we reiterate that the analysis is grossly simplified by using the rwa . for @xmath109
the system has two integrals of motion , the effective energy @xmath110 and the manley - rowe invariant @xmath111 .
we note that @xmath110 is not the energy of the oscillators .
rather it is an analog of the quasienergy of a periodically driven quantum system @xcite in the limit of a vanishing field .
the @xmath109-dynamics ( [ eq : energy_reduced ] ) is thus integrable @xcite .
it maps on the dynamics of a pendulum . in the variables @xmath112 and @xmath113 the equations of motion read @xmath114 where @xmath115 is given by eq .
( [ eq : energy_reduced ] ) in which @xmath116 and @xmath117 are expressed in terms of @xmath118 and @xmath119
. equations of motion ( [ eq : pendulum ] ) can be explicitly solved in terms of the jacobi elliptic functions .
the dimensionless period of the vibrations @xmath120 can be expressed in terms of the complete elliptic integral of the first kind . on the phase plane
@xmath121 the hamiltonian trajectories ( [ eq : pendulum ] ) have the familiar pattern of the sets of closed and open ( along the @xmath122-axis ) trajectories separated by a separatrix loop . on this loop
the period @xmath120 diverges .
the hamiltonian formulation allows us to describe the effect of weak dissipation and weak driving in a simple form .
away from the separatrix loop the dissipation is slow compared to the vibrations ( [ eq : pendulum ] ) and the motion can be described in terms of the period - averaged dynamical variables .
we denote period - averaged variables by the overline .
of special interest to us is the period - averaged squared amplitude of mode 1 , which is given by the scaled variable @xmath123 with @xmath124 given by the solution of eq .
( [ eq : pendulum ] ) for given @xmath125 .
an example of this function is shown in fig .
[ fig : i_boring ] .
the evolution of @xmath126 due to the dissipation is controlled by the evolution of the quasienergy @xmath127 and the invariant @xmath119 .
the latter is described by equation @xmath128 and a similar equation for @xmath119 .
one can easily see that the right - hand sides of eq .
( [ eq : evolution ] ) and the similar equation for @xmath119 are functions of @xmath127 and @xmath119 .
once they are found , one can find the time evolution of the scaled squared vibration amplitude @xmath126 from eq .
( [ eq : averaging_defined ] ) .
this is a theoretical basis for describing ring - down measurements for weakly damped modes that undergo nonlinear resonance .
the examples will be presented in a separate paper . in the presence of driving , the hamiltonian system with quasienergy @xmath129
is no longer integrable .
the analysis is simplified in the case of weak driving and weak dissipation . the driving pumps energy into the system , and the dissipation provides an energy drain .
they compensate each other in the stationary regime . when both are slow compared to the vibration period @xmath120 , the system reaches a regime of periodic vibrations with the values of @xmath125 determined by the balance equation @xmath130 , and a similar equation for @xmath119 .
in contrast to the dissipative term , the driving - induced change of the quasienergy @xmath131 has to be calculated to the second order in the @xmath132 and then averaged over the period .
an important immediate consequence of the presented analysis is that , for weak damping , the stationary regime of the coupled modes corresponds to vibrations of the mode amplitudes in the rotating frame .
these vibrations have the period @xmath133 , which is much larger than the driving period @xmath27 .
they correspond to sustained amplitude modulated responses in the driven system , which often appear via hopf bifurcations and can ultimately evolve into chaos @xcite .
the presented analysis describes the dynamics of systems with weakly damped vibrational modes that experience nonlinear resonance .
the specific results refer to 1:3 resonance
. the peculiar features of the vibrational dynamics come from the strong dependence of the coupling strength on the mode amplitude and the amplitude dependence of the mode frequencies , which leads to tuning the modes into and out of resonance as their amplitudes vary .
another factor is the interplay of the nonlinearity and the decay . the nonlinear resonant intermodal coupling results in non - exponential decay of the vibrations , with the decay rate that strongly depends on amplitude .
it also leads to a non - monontonic dependence of the effective vibration frequency on amplitude even where , in the absence of the modal coupling , this dependence is monotonic and comes from the simple duffing nonlinearity .
the decay rate and the frequency dependence on amplitude could be obtained in an explicit form in the important case where the high - frequency mode decays much faster than the low - frequency one , and thus the high - frequency mode adiabatically follows the low - frequency mode .
resonant driving in this case leads to coexistence of several stable states of forced vibrations .
the switching between the branches of the stable states with varying parameters of the drive has a peculiar pattern , where some branches can be reached by varying only the drive strength , but not the frequency , for example .
the analysis indicates that the stable states correspond to vibrations with time - independent amplitudes .
in contrast , if the decay rates of the modes are small compared to the nonlinearity , in the absence of the drive the dynamics in the rotating frame can be mapped onto a pendulum . with weak dissipation and weak drive
, the stable states correspond to nonsinusoidal oscillations of the amplitude .
the decay of the vibration amplitude after the drive is switched off is also profoundly nonexponential and can be accompanied by oscillations of the amplitude .
it is described in terms of the decay of the effective quasi - energy and the manley - rowe invariant .
we are grateful to adrian bachtold , changyao chen , dave czaplewski , andreas isacsson , daniel lpez , pavel polunin , and scott strachan for fruitful discussions .
this work was supported in part by the us army research office ( w911nf-12 - 1 - 0235 ) and the us defense advanced research projects agency ( fa8650 - 13 - 1 - 7301 ) .
s.w.s acknowledges partial support from the natioinal science foundation ( grant no.cmmi-sdc 1561934 ) .
o.s . and s.w.s were supported in part by florida institute of technology .
m. i. d. acknowledges partial support from the national science foundation ( grant no .
dmr-1514591 ) . )
32ifxundefined [ 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 _
( , ) link:\doibase 10.1007/bf01341712 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop _ _ ( , ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1103/physrevlett.110.127202 [ * * , ( ) ] @noop * * , ( ) @noop * * ( ) @noop * * , ( ) @noop ( ) , @noop ( ) , @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1126/science.aaa2085 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1021/nl401978p [ * * , ( ) ] link:\doibase 10.1021/nl403936a [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1103/physreva.7.2203 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) | we study nonlinear resonance of coupled modes in nano - mechanical systems . to reveal the qualitative features of the dynamics
, we consider the limiting cases , where the results can be obtained analytically . for 1:3 resonance ,
we find the anomalously strong and nonmonotonic dependence of the decay rate of the low frequency mode on its amplitude , if the decay rate of the high - frequency mode is comparatively large . in this case
the low - frequency mode driven close to resonance can have several branches of steady - state vibrations with constant amplitude .
if the decay rates of the both modes are small compared to their coupling and internal nonlinearity , the dynamics corresponds to slowly decaying strongly nonsinusoidal oscillations of the vibration amplitude .
weak driving can make these vibrations stable . |
experiments with radioactive ion beams ( rib ) started a new era in the nuclear reaction physics for the last decades @xcite .
with regard to probing and understanding the nuclear structure , in these experiments , some unexpected properties of the nucleus have been discovered .
one of the most intriguing attributes is the halo structure @xcite , consisting of a core and a weakly - bound valance nucleon(s ) to this core . up to now , this phenomenon has been greatly investigated experimentally on the various targets @xcite and caused a challenge for nuclear reaction theoreticians to reproduce the experimental data @xcite .
@xmath0 is one of the four one - neutron halo nuclei together with @xmath3 @xcite , and newly reported @xmath4 @xcite , and @xmath5 @xcite .
some experiments has been conducted for understanding of structure of @xmath0 .
firstly , tanihata et .
al @xcite observed a large radii for @xmath0 compared to @xmath2 in cross section measurements with targets at 90 mev / a , and concluded the halo structure for @xmath0 originating from its small neutron separation energy of 0.503 mev .
a few years later , fukuda et al .
@xcite confirmed this conclusion in elastic scattering of @xmath0 on c and al targets at 33 mev / nucleon . since these distinguished works
, @xmath0 has been still studied experimentally @xcite and theoretically @xcite .
one experimental study about @xmath0 is performed by cortina - gil et al .
@xcite for the cross section of the elastic scattering on @xmath1 at 49.3 mev / nucleon incident energy .
the first theoretical investigation of this measurement is an adiabatic approximation assuming no internal motion between the valance nucleon and the core in projectile , and also neglecting the interactions between the valance nucleon and the target nucleus @xcite .
also , in the same year , al - khalili et al .
@xcite investigated the same reaction with the few - body glauber model , in which the particles of the projectile are considered as following straight line paths through the interaction field of the target .
in addition to these studies , the continuum - discretized coupled - channels ( cdcc ) method was applied to this elastic scattering by takashina et .
@xcite , and also they used the same parameter set for the optical potentials between the projectile components and the target as in ref .
@xcite . in this non - adiabatic method , due to the very low neutron or proton separation energy , the continuum states of the projectile above this threshold energy
are discretized to finite number of states using momentum bins . including the breakup effects into the theoretical calculations of mentioned methods
gives almost the same results .
in the present study , the elastic scattering of the projectile halo nucleus @xmath0 on the target @xmath1 at 49.3 mev / nucleon @xcite is investigated as a breakup reaction using the optical model with the aid of a nuclear structure model .
different from other studies , the optical model potential used for the interaction between the core nucleus @xmath2 and the target @xmath1 is obtained by fitting to elastic cross section data at 39.1 mev / nucleon .
this data @xcite is investigated with optical model for the first time in this study . as for the interaction between the valance neutron of the halo nucleus and the target ,
the optical potential is deduced from an interpolation for different incident energies of neutron on @xmath1 target by means of global potential of ref .
@xcite . in order to describe non - elastic contributions
, we use a surface potential , which can be named as dpp ( dynamical polarization potential ) or vcp ( virtual coupling potential ) , in our calculations .
finally , a binding potential is employed for n+@xmath2 system .
unlike similar studies , we determine the value of the spectroscopic factor , describing the wave function of @xmath0 in terms of the wave function of @xmath2 , with the method given in ref .
@xcite for the direct neutron capture reaction @xmath2+@xmath6+@xmath7
. however , we use a new nuclear level density ( nld ) model @xcite , which strongly depends on the deformation of nucleus .
this paper organized as follows : the method used in this study is presented in sec .
[ theory ] , the results obtained by this method are given for the breakup reaction of @xmath0 in sec . [ results ] , and finally in sec . [ conclusions ] , concluding remarks drawn from this study are given .
since the mid - fifties , the optical model has been greatly used to investigate the elastic scattering cross section for both light and heavy ions in a wide range of incident energy .
the optical model considers the projectile and the target nuclei as the structureless particles in order to avoid many - body problems in nuclear physics calculations , and describes the interaction between the projectile and target with an effective potential . in this work , since we included breakup effects , the halo projectile @xmath0 is considered as a two - body system , which consists of @xmath2 core and a valance neutron . therefore , we define the effective potentials between projectile components and the target @xmath1 , which are n+@xmath1 , @xmath2+@xmath1 and n+@xmath2 as used ref .
@xcite @xmath8 where c , t , v correspond to @xmath2 core , @xmath1 target and valance nucleon , respectively .
an effective potential is a combination of the following terms as @xmath9 the first term is the centrifugal potential , which is defined as @xmath10 traditionally .
uniformly charged sphere assumption is employed for nucleus , @xmath11 where the charge radius @xmath12 is defined as @xmath13 , the coulomb potential parameter @xmath14 is taken as 1.2 fm in this work . in the optical model ,
the volume term in a effective potential has a crucial role and can be described with the real part of this term .
however , for inelastic contributions , an imaginary part is added to the volume term for the purpose of considering absorption of the reaction flux from the elastic channel to inelastic reaction channels .
therefore , conventionally the volume term consists of real and imaginary parts in reaction studies , @xmath15 where potential depths , radius and surface diffuseness parameters for both real and imaginary parts should be adjusted elastic scattering data . even if the investigated reaction is the elastic scattering , non - elastic contributions can still exist in the elastic channels . to include these contributions ,
the surface potential are used @xmath16 which is in derivative form of the volume term . final term in eq .
is the spin - orbit potential @xmath17 where @xmath18 .
the optical potential parameters in these equations can be determined from elastic scattering data . as a first step in fitting procedure of potential parameters ,
the geometrical parameters are adjusted to positions of peaks occurred in data .
afterwards , the potential depths of all used optical model potentials are fitted to experimental data to give the minimum @xmath19 value . in the case of halo nucleus
@xmath0 , the spectroscopic factor as a structure property is used to describe the ground state and the first excited state of @xmath0 in terms of @xmath2 .
the spectroscopic factor can be determined from the fitting to experimental cross section data of transfer or direct capture processes , and also they can be obtained theoretically from the shell model calculations . in literature , many transfer processes include the spectroscopic factor value of @xmath0 for @xmath20(t , p)@xmath0 @xcite , @xmath2(d , p)@xmath0 @xcite and @xmath0(p , d)@xmath2 @xcite reactions
. however , the experimental data of the direct capture cross section for @xmath2(n,@xmath7)@xmath0 is not available , but the direct capture cross section data can be deduced from coulomb dissociation @xcite . as a tool for calculations of the light ion cross sections such as direct capture processes , the nuclear level density has a crucial role to reproduce the measured data and define the spectroscopic factor .
therefore , the relation between the direct capture cross section and the nuclear level density , which is the number of the excited levels around an excitation energy , can be defined as @xcite @xmath21 where s represents the average spectroscopic factor , and @xmath22 is the level density function in terms of the excitation energy @xmath23 , total angular momentum @xmath24 , and the parity @xmath25 of compound nucleus . in the present work ,
we calculate the direct neutron capture cross section and compare to deduced data @xcite from coulomb dissociation of @xmath0 measured by nakamura et .
al . @xcite . to do this calculation ,
laplace - like formula @xcite is used for the energy dependence of the nuclear level density parameter in fermi gas model . according to this formula the level density parameter
strongly depends on the deformation of the nucleus , and the results obtained with this formula are very successful to describe low - lying collective levels compared to other phenomenological level density models @xcite .
therefore , keeping in mind that @xmath2 and @xmath0 are well - deformed nuclei , we expect that this formula is convenient to explain the neutron capture cross section data of @xmath2 . in the following section
, we will give the optical potential parameters , which are used in this study , and the results of our calculations .
to describe the interactions between the projectile and target , we consider the weakly - bound nucleus @xmath0 as @xmath2+n on @xmath1 target . for this purpose , firstly we focus on the interaction between the neutron and the target
. great number of experimental data in 0 - 100 mev energy range @xcite is found for the elastic scattering of neutron on @xmath1 , and can be used to define the effective potential between the valance nucleon and the target for in case @xmath0 .
unfortunately , for 49.3 mev incident energy , no experimental data is available .
thus , an interpolation of the global parametrization @xcite is used .
the results obtained with this global potential are given in figure [ fig1 ] .
as seen from figure , this interpolation of the global parametrization for n+@xmath1 at 49.3 mev incident energy gives us a good agreement in wide range energy .
( color online ) cross sections for n+@xmath1 at 28.2 mev @xcite , 35.0 mev @xcite , 65.0 mev @xcite , 75.0 mev @xcite , 85.0 mev @xcite , 94.8 mev @xcite , 95.0 mev @xcite , 96.0 mev @xcite . obtained results using the optical potentials for 49.3 mev incident energy
are represented by black line . ] in contrast to @xmath0 , very long - lived ( @xmath26y ) and a tightly - bound nucleus @xmath2 has a greater neutron separation energy , which is 6.81 mev .
one of the experimental study about @xmath2 is lapoux et .
@xcite in which they measured the elastic cross section for @xmath2 and @xmath0 projectiles on proton and @xmath1 targets at 39.1 mev / nucleon and 38.4 mev / nucleon , respectively , and these data was investigated using the microscopic jeukenne - lejeune - mahaux nucleon - nucleus potential for proton target and framework of folding model for c target .
unlike the other studies , @xcite in order to be more physical and reliable , the potential parameters describing the interaction between the core and the target are adjusted to the elastic scattering data at 39.1 mev / nucleon energy @xcite .
our obtained values of the potential depth parameters are shallow compared to their potential .
we use the experimental @xmath27 quadrupole deformation value , which is -0.6 @xcite , for the first @xmath28 excited level of @xmath1 , which is 4.4 mev .
also , in order to take into account the non - elastic contributions caused from the interactions at surface region , additionally one can add the surface term into the effective potential .
this potential is sometimes referred as surface term , derivative of woods - saxon potential form , dpp and vcp , and can be obtained by different methods .
the parameters of dpp can be obtained from microscopical @xcite or phenomenological @xcite calculations by fitting to the experimental data . for @xmath2+@xmath1
, we used a phenomenological dpp obtained from the fit process to the experimental data combined with a volume term .
obtained results for this elastic scattering and the optical potential parameters used in this calculation are given in [ fig2 ] and table [ tab01 ] , respectively .
except from the well - known phenomenon at @xmath29 , the data is reproduced well . llccccc interaction & type & @xmath30(mev ) & @xmath31(fm ) & @xmath32(fm ) + potential & & @xmath33(mev ) & @xmath34(fm ) & @xmath35(fm ) + + [ -1.8ex ] n+@xmath1 & volume&37.5 & 1.127&0.676 & + & & 4.90 & 1.127&0.676 & + & surface & 0.00 & 1.306&0.543 & + & & 4.15 & 1.306&0.543 & + & spin - orbit & 4.68 & 0.903&0.590 & + & & -0.39&0.903&0.590 & + + [ -1.8ex ] @xmath2+@xmath1 & volume&15.049 & 0.950&0.580 & + & & 23.326 & 1.100&0.630 & + + [ -1.8ex ] n+@xmath2 & volume&37.5 & 1.127&0.676 & + + [ -1.8ex ] @xmath0+@xmath1 & surface&42.793 & 0.950&0.580 & + sf=1.48 & & 3.935 & 1.100&0.630 & + + [ -1.8ex ] @xmath0+@xmath1 & surface&29.635 & 1.100&0.580 & + sf=0.71,0.62 & & 1.036 & 1.100&0.630 & + ( color online ) cross sections for @xmath2+@xmath1 target at 39.1 mev .
solid red line represents the results obtained by using the optical potential parameters given in table [ tab01 ] .
the experimental data is taken from @xcite . ]
many authors analysed the elastic scattering of the halo - nuclues @xmath0 on target @xmath1 for this incident energy by different theoretical models @xcite .
however , none of these studies is incorporated the nuclear structure to explain the data . on the other hand , adding the nuclear structural information into the reaction calculations for such a weakly - bound system as the halo nucleus is played a crucial role to contribute the agreement between the predictions and the experimental data .
the spectroscopic factor as a nuclear structure property is one of the most important ingredients for the theoretical cross section calculations of both light and heavy ions .
there are many methods which can be used for determining the value of the spectroscopic factor .
of course , the easiest method is to fit the spectroscopic factor values to the experimental cross section data directly , but the traditional way of estimating spectroscopic factor is to use the shell model in which the spectroscopic factor is defined as the square of normalization of the overlap integral between the wave function of the valance nucleon in the state of the target nucleus and the residual nucleus . also , the spectroscopic factor is a key ingredient for the direct capture process for which the related cross section often dominates the total cross section at the very low energies of astrophysical interest .
the direct capture process is can be used for obtaining the spectroscopic factor and known to play a notable role for light exotic nuclei systems for which few , or even no resonant states are available .
although many works containing the spectroscopic factors derived from the transfer processes are existing for the halo nucleus @xmath0 , the direct neutron capture cross section data for @xmath2 to compose @xmath0 is not available in the literature .
however , the direct capture cross sections can be obtained deduced data from the coulomb dissociation .
( color online ) the direct neutron capture cross section results for @xmath2(n+@xmath7)@xmath0 reaction at 0-3.mev lab . energy .
the solid red line represents the results of the present work using the level density model @xcite , and obtained spectroscopic factor value is 1.48 .
the deduced experimental data from coulomb dissociation data of @xmath0 @xcite is taken from ref .
] in obtaining the spectroscopic factor with the aid of the direct capture cross section calculations , the most important component is the nuclear level density .
generally , the reasons for not trusting to level density models to use in such calculations are their insufficient agreements with the experimental observables and their way of taking into account the collective effects . for overcoming these challenges , recently , we introduced a new laplace - like formula @xcite for the nld parameter to improve the predictive power for describing the low - lying collective levels , which are well known to be of vital importance for the direct capture process . with this formula ,
a great agreement is achieved with the experimental observables .
therefore , the direct neutron capture cross section calculation based on this level density model for @xmath2(n+@xmath7)@xmath0 processes is shown in figure [ fig3 ] . although the data could not be reproduced below 0.5 mev , in the rest of the energy range the same behaviour is well explained .
the obtained average value for the spectroscopic factor is 1.48 .
the value of parameters used in our level density calculation are 1.345 for @xmath36 and 0.285 for @xmath37 , which are taken from our previous study .
( color online ) the elastic cross section results for @xmath0+@xmath1 at 43.9 mev .
the dashed blue line is the breakup calculation with the spectroscopic factors 0.71 for the ground state and 0.62 for the first excited state @xcite .
the solid red line represents the cross section result with the spectroscopic factor value of 1.48 obtained from the direct capture cross section .
the experimental data is taken from @xcite . ] considering @xmath0 as a two - body projectile , all values of the optical potential parameters are given in table [ tab01 ] for n+@xmath1 , n+@xmath2 and @xmath2+@xmath1 interactions .
the parameter values of potential depths for @xmath2+@xmath1 at 39.1 mev incident energy are rearranged as 46.3 mev and 13.8 mev of real and imaginary parts , respectively .
the same procedure is repeated for the surface potential as 9.820 mev and 3.661 mev . also , to include non - elastic contribution for @xmath0+@xmath1 , a surface potential
is added to effective potential . moreover , to compare our results , we perform another calculation with the spectroscopic factor obtained by schmitt et al .
@xcite which is 0.71 for the ground state and 0.62 for the first excited state , respectively . the results of this calculation is also shown in figure [ fig4 ] with dashed blue line . in our calculations
the average value of the spectroscopic factor is taken as the spectroscopic factor of the ground state .
since the spectroscopic factor of the first excited state has less effect on the results , the value of this factor is taken as 1.0 .
finally , our prediction for the elastic scattering cross section of @xmath0 on @xmath1 is shown in figure [ fig4 ] with solid red line .
as seen from this figure , the inclusion of the nuclear level density with the laplace - like formula in the reaction calculations has a very well effect on reproducing the cross section data .
also , the fit method we used for the optical potential parameters , which is to adjust the geometrical parameters to positions of peaks and the depths to give minimum @xmath19 , effects the agreement in a positive way .
in summary , we have investigated the elastic scattering cross section data of the projectile @xmath0 on @xmath1 target at 49.3 mev / nucleon @xcite . to include breakup effects into the calculations ,
the different optical potentials for n+@xmath1 , n+@xmath2 and @xmath2+@xmath1 are described .
also , the present study contains the first analysis of the phenomenological optical model for 39.1 mev incident energy of the projectile @xmath2 on @xmath1 target .
obtained results are in better agreement compared to the microscopic study of lapoux et al @xcite , which is the first and the only study of this reaction .
not only the effect of including the spectroscopic factor into calculations are found significant for the breakup reaction of @xmath0 but also adjusting the geometrical parameters to positions of peaks and the depths to give minimum @xmath19 gives positives contributions to reproduce the scattering data . the theoretical framework used for obtaining the spectroscopic factor by using the nuclear nuclear level density to calculate the direct neutron capture cross section
is employed for the first time in breakup reaction calculation of @xmath0 .
moreover , the nuclear level density is used for the first time as a spectroscopic tool in a light exotic nuclei induced reaction .
consequently , beside the success of the nuclear level density with the laplace - like formula for the level density parameter @xcite as a structure model , the results show that this new method seems appropriate to perform the reaction calculations .
99 i. tanihata , h. hamagaki , o. hashimoto , y. shida , n. yoshikawa , k. sugimoto , o. yamakawa , t. kobayashi , n. takahashi , phys .
lett . * 55 * , 2676 ( 1985 ) .
i. tanihata , h. hamagaki , o. hashimoto , s. nagamiya , y. shida , n. yoshikawa , o. yamakawa , k. sugimoto , t. kobayashi , d.e .
greiner , t. takahashi , y. nojiri , phys .
b 160 * , 380 ( 1985 ) . i. tanihata , t. kobayashi , o. yamakawa , s. shimoura , k. ekuni , k. sugimoto , n. takahashi , t. shimoda , h. sato , phys . lett .
* b 206 * , 592 ( 1988 ) .
k. c. c. pires , r. lichtenthaler , a. lepine - szily , v. guimaraes , p. n. de faria , a. barioni , d. r. mendes junior , v. morcelle , r. pampa condori , m. c. morais , j. c. zamora , and e. crema , a. m. moro , m. rodriguez - gallardo , m. assuncao , j. m. b. shorto , s. mukherjee , phys . rev . * c 83 * , 064603 ( 2011 ) .
kolata , m. zahar , r. smith , k. lamkin , m. belbot , r. tighe , b.m .
sherrill , n.a .
orr , j.s .
winfield , j.a .
winger , s.j .
yennello , g.r .
satchler , a.h .
wuosmaa , phys .
* 69 * , 2631 ( 1992 ) .
cortina - gil , p.
roussel - chomaz , n. alamanos , j. barrette , w. mittig , f.s .
dietrich , f. auger , y. blumenfeld , j.m .
casandjian , m. chartier , v. fekou - youmbi , b. fernandez , n. frascaria , a. gillibert , h. laurent , a. lepine - szily , n.a .
orr , j.a .
scarpaci , j.l .
sida , t. suomijarvi , phys . lett .
* b 401 * , 9 ( 1997 ) .
i.j . thompson , j.s .
al - khalili , j.a .
tostevin , j.m .
bang , phys . rev .
* c 47 * , r1364 ( 1993 ) .
johnson , j.s .
al - khalili , j.a .
tostevin , phys .
lett . * 79 * , 2771 ( 1997 ) .
t. nakamura , n. fukuda , t. kobayashi , n. aoi , h. iwasaki , t. kubo , a. mengoni , m. notani , h. otsu , h. sakurai , s. shimoura , t. teranishi , y. x. watanabe , k. yoneda , m. ishihara , phys .
lett . * 83 * , 1112 ( 1999 ) .
t. nakamura , n. kobayashi , y. kondo , y. satou , j. a. tostevin , y. utsuno , n. aoi , h. baba , n. fukuda , j. gibelin , n. inabe , m. ishihara , d. kameda , t. kubo , t. motobayashi , t. ohnishi , n. a. orr , h. otsu , t. otsuka , h. sakurai , t. sumikama , h. takeda , e. takeshita , m. takechi , s. takeuchi , y. togano , k. yoneda , phys .
* 112 * , 142501 ( 2014 ) .
n. kobayashi , t. nakamura , y. kondo , j. a. tostevin , y. utsuno , n. aoi , h. baba , r. barthelemy , m. a. famiano , n. fukuda , n. inabe , m. ishihara , r. kanungo , s. kim , t. kubo , g. s. lee , h. s. lee , m. matsushita , t. motobayashi , t. ohnishi , n. a. orr , h. otsu , t. otsuka , t. sako , h. sakurai , y. satou , t. sumikama , h. takeda , s. takeuchi , r. tanaka , y. togano , k. yoneda , phys .
lett . * 112 * , 242501 ( 2014 ) .
m. fukuda , t. ichihara , n. inabe , t. kubo , h. kumagai , t. nakagawa , y. yano , i. tanihata , m. adachi , k. asahi , m. kouguchi , m. ishihara , h. sagawa , s. shimoura , phys .
* b 268 * , 339 ( 1991 ) . v. lapoux , n. alamanos , f. auger , y. blumenfeld , j .-
casandjian , m. chartier , m.d .
cortina - gil , v. fekou - youmbi , a. gillibert , m. mac cormick , f. marechal , f. marie , w. mittig , f. de oliveira santos , n.a .
orr , a.n .
ostrowski , s. ottini - hustache , p. roussel - chomaz , j .- a .
scarpaci , j .- l .
sida , t. suomijarvi , j.s .
winfield , phys .
* b 658 * , 198 ( 2008 ) .
a. di pietro , g. randisi , v. scuderi , l. acosta , f. amorini , m.j.g .
borge , p. figuera , m. fisichella , l.m .
fraile , j. gomez - camacho , h. jeppesen , m. lattuada , i. martel , m. milin , a. musumarra , m. papa , m.g .
pellegriti , f. perez - bernal , r. raabe , f. rizzo , d. santonocito , g. scalia , o. tengblad , d. torresi , a. maira vidal , d. voulot , f. wenander , and m. zadro , phys .
lett . * 105 * , 022701 ( 2010 ) .
r. anne , s.e .
arnell , r. bimbot , s. dogny , h. emling , h. esbensen , d. guillemaud - mueller , p.g .
hansen , p. hornshj , f. humbert , b. jonson , m. keim , m. lewitowicz , p. mller , a.c .
mueller , r. neugart , t. nilsson , g. nyman , f. pougheon , k. riisager , m .-
saint - laurent , g. schrieder , o. sorlin , o. tengblad , k. wilhelmsen rolander , d. wolski , phys .
lett * b 304 * , 55 ( 1993 ) .
n. fukuda , t. nakamura , n. aoi , n. imai , m. ishihara , t. kobayashi , h. iwasaki , t. kubo , a. mengoni , m. notani , h. otsu , h. sakurai , s. shimoura , t. teranishi , y. x. watanabe , and k. yoneda , phys .
* c 70 * , 054606 ( 2004 ) .
summers , f.m .
nunes , i.j .
thompson , phys .
* c 73 * , 031603 ( 2006 ) .
howell , j.a .
tostevin , j.s .
al - khalili , j. phys .
g : nucl . part .
phys . * 31 * , s1881 ( 2005 ) .
dasso , s.m .
lenzi , a. vitturi , phys . rev . * c 59 * , 539 ( 1999 ) .
al - khalili , j.a .
tostevin , j.m .
brooke , phys . rev .
* c 55 * , r1018 ( 1997 ) .
m. takashina , s. takagi , y. sakuragi , y. iseri , phys .
* c 67 * , 037601 ( 2003 ) .
a.j . koning , j.p .
delaroche , nucl .
phys . * a 713 * , 231 ( 2003 ) .
s. goriely , astron
. astrophys . * 325 * , 414 ( 1997 ) .
b. canbula , r. bulur , d. canbula , h. babacan , nucl .
phys . * a 929 * , 54 ( 2014 ) .
summers , f.m .
nunes , i.j .
thompson , phys . rev .
* c 74 * , 014606 ( 2006 ) .
pullen , a.e litherland , s. hinds , r. middleton , nucl .
36 * , 1 ( 1962 ) .
robertson , e. kashy , w. benenson , a. ledebuhr , phys . rev .
* c 17 * , 4 ( 1978 ) .
f. azjenberg - selove , e.r .
flynn , o. hansen , phys .
* c 17 * , 1283 ( 1978 ) .
liu , h.t .
fortune , phys .
* c 42 * , 167 ( 1990 ) .
goosman , r.w .
kawanagh , phys . rev .
* c 1 * , 1939 ( 1970 ) .
auton , nucl . phys . * a 157 * , 305 ( 1970 ) .
b. zwieglisnki , w. benenson , r.g.h .
robertson , w.r .
coker , nucl .
phys . * a 315 * , 124 ( 1979 ) .
schmitt , k.l .
jones , s. ahn , d.w .
bardayan , a. bey , j.c .
blackmon , s.m .
brown , k.y .
chae , k.a .
chipps , j.a .
cizewski , k.i .
hahn , j.j .
kolata , r.l .
kozub , j.f .
liang , c. matei , m. matos , d. matyas , b. moazen , c.d .
nesaraja , f.m .
nunes , p.d .
omalley , s.d .
pain , w.a .
peters , s.t .
pittman , a. roberts , d. shapira , j.f .
shriner , jr .
smith , i. spassova , d.w .
stracener , n.j .
upadhyay , a.n .
villano , g.l .
wilson6 , phys .
rev * c 88 * , 064612 ( 2013 ) .
winfield , s. fortier , w.n .
catford , s. pita , n.a .
orr , j. van de wiele , y. blumenfeld , r. chapman , s.p.g .
chappell , n.m .
clarke , n. curtis , m. freer , s. gales , h. langevin - joliot , h. laurent , i. lhenry , j.m .
maison , p. roussel - chomaz , m. shawcross , k. spohr , t. suomijarvi , a. de vismes , nucl . phys . * a 683 * , 48 ( 2001 ) .
a. mengoni , t. otsuka , t. nakamura , m. ishihara , nucl . phys . * a 621 * , 323c ( 1997 ) . t. nakamura , s. shimoura , t. kobayashi , t. teranishi , k. abe , n. aoi , y. doki , m. fujimaki , n. inabe , n. iwasa , k. katori , t. kubo , h. okuno , t. suzuki , i. tanihata , y. watanabe , a. yoshida , m. ishihara , phys
* b 331 * , 296 ( 1994 ) .
a.j . koning , s. hilaire , s. goriely , nucl .
phys . * a 810 * , 13 ( 2008 ) .
y.yamanouti , m.sugimoto , s.chiba , m.mizumoto , y.watanabe , k.hasegawa , jaeri - m reports * 199 * , 177 ( 1989 ) .
t.niizeki , h.orihara , k.ishii , k.maeda , m.kabasawa , y.takahashi , k.miura , nucl .
methods in physics res . *
287 * , 455 ( 1990 ) .
osborne , f.p .
brady , j.l .
romero , j.l .
ullmann , d.s .
sorenson , a. ling , n.s.p .
king , r.c .
haight , j. rapaport , r.w .
finlay , e. bauge , j.p .
delaroche , a.j .
koning , phys .
rev * c 70 * , 054613 ( 2004 ) .
p. mermod , j. blomgren , c. johansson , a. ohrn , m. osterlund , s. pomp , b. bergenwall , j. klug , l. nilsson , n. olsson , u. tippawan , p. nadel - turonski , o. jonsson , a. prokofiev , p .- u .
renberg , y. maeda , h. sakai , a. tamii , k. amos , r. crespo , and a. moro , phys .
rev * c 74 * , 054002 ( 2006 ) .
j. klug , j. blomgren , a. ata , b. bergenwall , a. hildebrand , c. johansson , p. mermod , l. nilsson , s. pomp , u. tippawan , k. elmgren , n. olsson , o. jonsson , a. v. prokofiev , p .- u .
renberg , p. nadel - turonski , s. dangtip , p. phansuke , m. osterlund , c. le brun , j. f. lecolley , f.r .
lecolley , m. louvel , n. marie - noury , c. schweitzer , ph .
eudes , f. haddad , c. lebrun , a.j .
koning , x. ledoux , phys . rev .
* c 68 * , 064605 ( 2003 ) .
vermeer , m.t .
esat , j.a .
kuehner , r.h .
spear , phys .
* b 122 * , ( 1987 ) .
hussein , g.r .
satchler , nucl . phys . * a 567 * , 165 ( 1994 ) .
khoa , g.r .
satchler , w. von oertzen , phys . lett .
* b 358 * , 14 ( 1995 ) .
pacheco , n. vinh mau , nucl .
phys . * a 669 * , 135 ( 2000 ) .
l.f . canto , r. donangelo , m.s .
hussein , nucl .
phys . * a 529 * , 243 ( 1991 ) .
canto , r. donangelo , m.s .
hussein , m.p .
pato , nucl . phys . * a 542 * , 131 ( 1992 )
. k. yabana , y. ogawa , y. suzuki , phys . rev .
* c 45 * , 2909 ( 1992 ) . k. yabana , y. ogawa , y. suzuki , nucl . phys . * a 539 * , 295 ( 1992 ) .
n. takigawa , m. ueda , m. kuratani , h. sagawa , phys . lett .
* b 288 * , 244 ( 1992 ) .
al - khalili , nucl . phys . * a 581 * , 315 ( 1995 ) .
y. sakuragi , s. funada , y. hirabayashi , nucl .
phys . * a 588 * , 65c ( 1995 ) .
schmitt , k.l .
jones , a. bey , s.h .
ahn , d.w .
bardayan , j.c .
blackmon , s.m .
brown , k.y .
chae , k.a .
chipps , j.a .
cizewski , k.i .
hahn , j.j .
kolata , r.l .
kozub , j.f .
liang , c. matei , m. matos , d. matyas , b. moazen , c. nesaraja , f. m. nunes , p.d .
omalley , s.d .
pain , w.a .
peters , s.t .
pittman , a. roberts , d. shapira , j f. shriner , jr .
smith , i. spassova , d. w. stracener , a. n. villano , and g. l. wilson , phys .
lett . * 108 * , 192701 ( 2012 ) . | the experimental elastic cross section data of the projectile @xmath0 on target @xmath1 at 49.3 mev / nucleon energy is analysed .
the calculations for the elastic scattering is performed by the phenomenological optical model .
the different optical potentials to include breakup effects into the calculations , which are neutron+@xmath1 , neutron+@xmath2 and @xmath2+@xmath1 , are described with the aid of the global potentials for neutron interactions and fitted to experimental data for the core and target interaction . also , the first analysis of the optical model for @xmath2 on target @xmath1 at 39.1 mev is done for building the interaction potential of the core and the target for @xmath0 . for investigating the effects of the spectroscopic factor obtained from the direct capture process using the nuclear level density
are compared with the previous cross section and spectroscopic factor results .
obtained results for the elastic cross section are reproduced the experimental data very well , and shows the requirement of including spectroscopic properties such as the spectroscopic factors and the density of the excited states to explain this elastic cross section data . |
alnico exhibits quasicrystalline phases over a range of compositions and temperatures@xcite . of special interest
is the ni - rich quasicrystalline phase around the composition al@xmath0ni@xmath1co@xmath2 .
this is a decagonal phase with a period of 4.08 along the periodic axis , making it a simple phase relative to other members of the al - ni - co family .
additionally , it appears to be most perfect structurally , even though it is stable only at high temperatures around t=1000 - 1200k .
its structure has been extensively studied experimentally by x - ray diffraction@xcite and electron microscopy@xcite . finally , since qualitatively accurate pair potentials are available@xcite , structural predictions can be made based on total energy considerations@xcite .
an idealized deterministic structure for this phase has been proposed by studying the total energy@xcite .
this prediction employed a multi - scale simulation method , in which small system sizes were simulated starting with very limited experimental input , then rules discovered through the small scale simulations were imposed to accelerate simulations of larger - scale models .
although efficient , this approach leaves open the question of how strongly the final model was biased by the chosen method .
we adopt a different approach here , starting from slightly different experimental input and working directly at the relevant high temperatures .
the experimental input is : ( 1 ) the density @xcite , composition and temperature at which the phase exists ; ( 2 ) the hyperspace positions of atomic surfaces ( these are simply the positions for a penrose tiling , with as1 at @xmath4 and as2 at @xmath5 ) ; ( 3 ) the fact that the quasicrystal is layered , with space group @xmath6 , and its lattice constants ( we take @xmath7 , @xmath8 ) .
the chief unknowns to be determined are the sizes , shapes and chemical occupancies of the atomic surfaces .
like the prior study @xcite , we employ monte carlo simulation using the same electronic - structure derived pair potentials@xcite .
also , like the prior study , we employ a discrete list of allowed atomic positions . however , instead of using sparse decorations of fundamental tiles , where the configurational freedom arose largely from flipping of the decorated tiles , in the present study we employ a fixed site list based on a rich decoration of fixed tiles . for a given tiling
the density of allowed sites in our new simulation is much greater than the actual atomic density .
the resulting atomic surface ( see fig . [
fig : model - perp]a ) corresponds to that of a penrose tiling with tile edge length @xmath9 ( plus a few additional sites such as inside some fat rhombi ) .
these atomic surfaces include within them the atomic surfaces previously proposed on the basis of total energy calculations @xcite , and analysis of experimental data @xcite .
this larger atomic surface avoids any bias towards particular atomic surface shapes .
( a ) oversized atomic surface used in the present simulation .
as1 ( left ) at @xmath10 , as2 ( right ) at @xmath11 . for comparison the ideal model proposed in ref .
@xcite is inscribed inside .
color scheme : black = co , gray = ni , light - gray = al , white = partial al occupancy .
( b ) total occupancy of as1 ( left ) and as2 ( right ) at t=1160k.,width=576 ] the following section describes the site list and other simulation details more precisely
. then we present site occupancy in real space and on the atomic surfaces .
we represent atomic surface occupancies using fourier - bessel series , allowing us to isolate the five - fold symmetric component and filter sampling noise .
such a representation could prove fruitful in crystallographic structure determination as well .
@xcite pursued related goals .
we compare our results with a recent experimental study .
briefly , we find broad agreement with the sizes , shapes and chemical occupancies of the atomic surfaces .
a few specific points of disagreement can be addressed by relaxing our fixed site positions .
finally , we conclude with a summary of our results and outlook for future work .
we should perform simulations in the continuum , with no predefined site list to bias our results .
unfortunately , strong binding of al to transition metal atoms ( co and ni ) leads to phase separation into a cscl - type crystal plus regions of excess pure al .
this reflects a deficiency of our interatomic potentials , which should not be applied at a transition - metal concentration as high as the 50% found in cscl - type crystals . restricting atomic positions a set that does not include the cscl crystal inhibits phase separation .
rather than specify these sites in a complex manner whose details may fail to allow some especially favorable structure , we take a relatively neutral assignment that still allows quasicrystalline structures to form .
specifically we restrict the possible atomic positions to the vertices of a penrose lattice whose edge length is small compared to the spacing between atoms .
the possible atomic sites fill space with a density 2.55 times greater than the actual atomic density , so the majority of sites are empty in any particular configuration .
our prior study of prefered structures proposed an idealized arrangement of atoms on an hbs ( hexagon - boat - star ) tiling of edge length @xmath12 , as illustrated in fig .
[ fig : model - para ] .
our monte carlo simulation distributes atoms among these sites in a manner consistent with thermal equilibrium at a given temperature .
we focus on the temperature t=1160k , which lies within the stability range for the ni - rich decagonal phase @xcite .
we simulate a low phason strain quasicrystal approximant , with lattice parameters @xmath13 @xmath14 , containing 9776 lattice sites and holding 3838 atoms .
we initially decorated the structure randomly at the desired density and composition .
we then lowered the temperature in steps from an initial t=4000k , annealing at each temperature , until we reached our data collection temperature of t=1160k .
our data analysis differs from conventional approaches to quasicrystal atomic surface modeling .
it is common to model atomic surface occupancy by breaking the atomic surface up into polygonal domains piecewise constant occupancy in each domain .
this approach highlights the manner in which occupancy depends on the local real space environment , which changes discontinuously as a function in perpendicular space .
unfortunately such a piecewise continuous representation does not properly capture the effects of chemical disorder and phason fluctuations characteristic of a quasicrystal at high temperatures .
we show below that the mean occupancy statistics actually vary quite smoothly over the atomic surface . to represent smoothly varying atomic surface occupancy functions it is helpful to introduce a complete basis set of analytic functions and expand the occupancy in this basis .
owing to the axial symmetry of the atomic surfaces it is handy to introduce polar coordinates @xmath15 . the natural function basis set for polar coordinates
are combinations of cylindrical bessel functions multiplied by complex exponentials .
thus , on a given atomic surface we express the occupancy by chemical species @xmath16 as @xmath17 the @xmath18 index represents the angular mode frequency . for a given angular frequency @xmath18 ,
the sum over @xmath19 allows the representation of an arbitrary radial variation , with radial frequencies @xmath20 such that all bessel functions share a common zero outside our atomic surface . to analyze our data
we invert eq .
( [ eq : bessel ] ) to obtain the coefficients @xmath21 using orthonormality of the basis set .
we then filter out the components whose angular variation lacks 5-fold symmetry by restricting @xmath18 to multiples of 5 .
each monte carlo step consisted of many attempted jumps ( of up to @xmath22 in length ) per atom .
an attempted jump to an occupied site means an attempt to swap the two atoms .
we recorded the detailed atomic site occupancy for 9000 configurations separated by sufficiently many monte carlo steps that the configurations become uncorrelated .
phason fluctuations such as bowtie flips @xcite cause the center of gravity of the structure to drift in perpendicular space . note that the lattice sites do not move during a phason flip , only their atomic occupancy changes . a bowtie flip in our ideal model interchanges two nini pairs with two alco pairs . to avoid smearing caused by this drift , which could obscure small - scale atomic surface details ,
we recentered the perp space center of gravity for each configuration before projecting onto the fourier - bessel coefficients . without recentering
the atomic surface occupancy functions would be weaker and broader . as in our prior @xcite low temperature study ,
we find that the hbs tiling effectively describes the quasiperiodic layers , but the prefered chemical decoration of the tiles now becomes highly variable at certain sites . fig .
[ fig : model - perp](b ) displays the average occupancy distribution on each atomic surface .
the as1 occupancy is a combination of all three chemical species .
the as2 occupancy is virtually 100% al . the breakdown of as1 occupancy among al , ni and co is shown in fig .
[ fig : as1alnico ] .
a striking feature of all these occupancy plots is the smooth , continuous variation of occupancy .
it is clearly not appropriate to model these densities using piecewise constant functions with polygonal boundaries , even though five - fold symmetry is present .
ni and co are strongly mixed with very similar distributions , concentrated at the center of as1 . because the amount of ni present exceeds the amount of co , the co occupancy vanishes more quickly than the ni occupancy away from the center , while the ni occupancy remains more nearly constant . in the region where the tm occupancy drops off it
is largely replaced with al so the total occupancy remains close to 100% in this crossover region .
in parallel space atoms in the tm to al crossover region correspond to tile edges .
effects of this gradient from tm to al occupancy can be seen in fig .
[ fig : model - para ] ( right ) , near tiling vertices at which two or more edges join in a @xmath23 angle .
the pair of occupied sites on these tile edges show unequal al / tm occupancy . the site closer to the vertex contains relatively less tm and more al than the site further away
the site closer to the vertex in parallel space lies in the atomic surface region that is further from the center in perp space .
note also that bowtie phason flips that replace tm
tm pairs with al tm pairs on tile edges are one type of excitation that replace the tm site close to a 72@xmath24 vertex with an al atom .
we understand these effects energetically because the inner edge site is in a tm - rich environment .
tm bonds are strongly favorable so this is a likely site for al occupancy .
far from the atomic surface centers , where the occupancy falls smoothly to zero , the atoms are mainly al .
these highly mobile atoms are in locations in parallel space where there are multiple ideal sites , too close to be simultaneously occupied , but each one with a rather similar local environment so that there is no strong energetic preference among the sites , for example the pair of symmetry - related sites near the center of the hexagon tile .
other examples are rings of 10 sites surrounding the ideal co position at the centers of boat and star tiles .
these rings can hold at most three al atoms .
jumps of al atoms among these sites correspond to phason tile flips of the underlying very small 1.52 rhombus tiling .
we compare primarily with the refinement by takakura @xcite .
we first point areas of general agreement between our results and the experimental data , then discuss the main points of disagreement .
our simulated occupation probabilities generally agree well with experiment .
the approximate sizes and shapes of our atomic surfaces agree with takakura .
we agree that as1 contains transition metal atoms within a central region ( takakura orbit numbers 1 - 5 ) , separated by a fully occupied region of mixed al / tm ( orbits 6 and 7 ) and finally radial spokes in which al occupancy diminishes from full to partial ( orbits 8 - 10 ) .
we agree that as2 ( orbit numbers 11 - 23 ) contains primarily al atoms .
takakura finds full occupancy of orbit numbers 14 and 16 , but only 82% occupancy of orbit numbers 15 and 17 , which are locally fairly similar ( these are the al sites adjacent to tile edge tm atoms ) .
our results qualitatively support this , with occupancy about 90% corresponding to orbit numbers 14 and 16 , dropping to around 70 - 80% for orbit numbers 15 and 17 . now consider discrepancies between our results and experiment .
orbits 19 and 23 , localized around special points on the fringes of as2 are fully occupied in the refinement , while we find partial occupancy . in parallel space
these correspond to pairs of ideal sites midway between tm atoms .
partial occupancy is forced because the spatial separation of these ideal points ( 2.25 ) is too small to allow both to be simultaneously occupied . however ,
if we consider multilayer strucures , with @xmath25 , and allow structural relaxation , we know @xcite it is possible to occupy 3 out of every 4 such sites per 8.16 .
the lack of structural relaxation thus causes an error in our occupancy for this orbit .
takakura s fully occupied orbits numbers 20 - 22 are located on the corners of as2 .
we find 60 - 80% occupancy instead of full occupancy , consistently with our smoothly decreasing occupation probability .
takakura also finds a small fraction ( about 20% ) of tm atoms on orbits 20 and 21 , while we find only al .
there is no strong energetic reason that tm atoms should not occupy those sites .
perhaps the tm occupancy of orbit number 20 is related to the large atomic displacement of this orbit takakura found during the refinement .
surprisingly , takakura finds 50% occupancy on orbit number 11 , at the very center of as2 . in parallel space
this site corresponds to hbs tile vertices at which five edges meet .
we find it is fully occupied , similar to all other hbs tile vertices .
we presented a method to compute atomic surface occupancy statistics , and applied it to the basic ni - rich decagonal phase of alnico .
this work launches a promising set of possible future research directions .
the method can be applied to other decagonal phases and also to icosahedral phases .
the main limitation is the availability of interatomic potentials .
further , a great deal remains to be done even within alnico , as we now discuss .
our simulations generate a complete set of equilibrium configurations , in which the precise location of every atom is known .
the analysis above only presents the atomic surface occupation probabilities , but does not address correlations among the partially occupied sites .
such correlations are contained in the patterson function , which is also known experimentally @xcite and can be computed from our existing data .
actual atomic configurations contain more information than the atomic surface occupancies , because all correlations are reflected in the actual configurations .
[ fig : model - para ] ( center ) illustrates an atomic arrangement from our simulation .
comparing with the ideal model ( left ) reveals a great deal of disorder , including phason flips , chemical disorder and vacancies . among the prominent correlations
are strong anticorrelation of too - close sites .
the deep clefts between arms of as1 correspond to sites that are close in parallel space to sites near the corners of as2 .
when these as2 sites are occupied sites within the clefts of as1 are empty .
when these as2 sites are empty then we find partial occupancy inside the clefts of as1 .
another anticorrelation occurs because the pairs of ideal sites midway between pairs of tm atoms ( e.g. takakura orbit number 19 and 23 ) are too close for simultaneous occupancy .
this anticorrelation arises because we used only `` ideal '' positions projected into parallel space from flat atomic surfaces . in reality
the atomic positions are displaced somewhat away from the ideal sites , and we showed above that such displacements can have a substantial influence on occupancy statistics in cases where the ideal sites are just slightly too close together .
indeed , relaxations of our ideal structures , using either pair potentials @xcite or full _ ab - initio _ calculations result in displacements similar to those reported experimentally @xcite . finally , our simulation used only a single 4.08 repeat unit along the 10-fold symmetry axis . owing to our periodic boundary conditions , swapping
a pair of atoms in the single repeat implies a simultaneous swap of all pairs of atoms in infinite columns extending above and below the original pair in our single repeat .
but when we calculate the energy for the swap , we only include the energies of the original atoms .
this type of finite - size error needs to be eliminated , as it miscalculates the energies of phason flips . simply stacking two repeats above each other to make an 8.16 unit , however , usually overestimates the energy cost of atom swaps corresponding to bowtie phason flips .
this is because uncorrelated flips between repeat units create phason stacking faults whose energy cost is large in the absence of atomic relaxation but is small , and can even be negative , when relaxation is included .
we must include relaxed atomic positions to achieve a realistic model of three - dimensional phason disorder .
we acknowledge discussions with c.l .
henley support by nsf grant dmr-0111198 .
0 s. ritsch , c. beeli _ et al . _ , phil . mag .
lett , * 78 * , 67 ( 1998 ) .
h. takakura , a. yamamoto and a .-
tsai , acta cryst .
a , * 57 * , 576 ( 2001 ) .
a. cervellino , t. haibach and w. steurer , acta cryst .
b , * 58 * , 8 ( 2002 ) .
e. abe , k. saitoh _
et al . _ ,
lett . , * 84 * , 4609 ( 2000 ) .
y. yan , s. j. pennycook and a .-
tsai , phys .
lett . , * 81 * , 5145 ( 1998 ) .
i. al - lehyani _ et al .
_ , phys . rev .
b , * 64 * , 075109 ( 2001 ) .
m. mihalkovi _
et al . _ ,
b , * 65 * , 104205 ( 2002 ) . l. elcoro and j. m. perez - mato and g. madaraiga , acta cryst .
a , * 50 * , 182 ( 1994 ) . m. v. jaric and s .- y .
qiu , phys .
b , * 49 * , 6614 ( 1994 ) . c. l. henley ,
m. mihalkovi and m. widom , j. alloys and comp .
, * 342 * , 221 ( 2002 ) .
w. steurer , t. haibach , b. zhang , s. kek and r. luck , acta cryst .
b , * 49 * , 66 , ( 1993 ) . | we investigate the high temperature decagonal quasicrystalline phase of al@xmath0ni@xmath1co@xmath2 using a quasilattice gas monte - carlo simulation . to avoid biasing towards a specific model
we use an over - dense site list with a large fraction of free sites , permitting the simulation to explore an extended region of perpendicular space . representing the atomic surface occupancy in a basis of harmonic functions
directly reveals the 5-fold symmetric component of our data .
occupancy is examined in physical and perpendicular space .
* equilibrium structure of decagonal alnico * s. naidu , m. mihalkovi@xmath3 and m. widom , + department of physics , carnegie mellon university , pittsburgh , pa 15213 + @xmath3also at : institute of physics , slovak academy of sciences , 84228 bratislava , slovakia 0.3 in |
dimer - based antiferromagnets ( dafs ) under a magnetic field are promising candidates for displaying new phases of bosonic matter.@xcite magnetic excitations in such systems , termed _ triplons _ , are well described by lattice models of interacting bosons , whose density can be finely tuned by varying the magnitude of the applied field.@xcite experimentally , field - induced bose - einstein condensation ( bec ) of triplons has been observed in a number of dafs ( see the review ref . ) and , remarkably , exotic quantum criticality has been detected in the spin - dimer compound @xmath0 .
@xcite the presence of magnetic frustration further adds to the rich phenomenology of these systems by enhancing repulsive interactions between triplons , something that may eventually stabilize incompressible phases that break the lattice s translational symmetry.@xcite such _ crystalline _ phases are for instance realized in the shastry - sutherland material @xmath1,@xcite where they are signaled by a series of magnetization plateaux at unconventional fillings stabilized by complex triplon interactions .
@xcite the occurrence of both bec and solid phases in the phase diagram of dafs under magnetic field suggests that the magnetic equivalent of the phase simultaneously displaying diagonal and off - diagonal order known as _
( ss)@xcite may be realized in these systems .
indeed , insofar as more exotic possibilities are excluded,@xcite according to the ginzburg - landau - wilson paradigm a continuous transition between phases breaking different symmetries ( as it is the case with bec and crystalline phases ) is precluded and we are therefore left with two possibilities : ( i ) a first - order transition or ( ii ) the appearance of an intermediate phase , where both order parameters coexist , termed spin - supersolid ( spin - ss ) in the present context .
the latter possibility has been first investigated by momoi and totsuka for the shastry - sutherland model in the vicinity of half- and third - filling plateaux,@xcite based on a mean - field analysis of an effective bosonic model derived up to third - order in the inter - dimer coupling .
more recently,@xcite state - of - the - art techniques have been employed in deriving effective models that improve upon the third - order effective hamiltonian of ref . .
unfortunately , the reliability of these methods is still limited to inter - dimer couplings equal to , at most , one - half of the intra - dimer coupling , and in that parameter range the different plateaux seem to be separated by first - order transitions without any convincing evidence of spin - ss phases.@xcite the situation is much clearer for the daf investigated in ref . where repulsion among triplons are enhanced due to the strong ising - like character of the inter - dimer exchange [ see eq .
( 1 ) in ref . ] , making room for checkerboard solid ( cbs ) and spin - ss phases to emerge .
the absence of frustration allows for quantum monte carlo ( qmc ) simulations to be performed and , in this way , the occurrence of a spin - ss phase for the model studied in ref . has been firmly established .
however , such a strongly anisotropic hamiltonian is unrealistic for mott insulating materials and further investigations of models where the kinetic energy is instead reduced by frustration of isotropic couplings@xcite are clearly called for if connection to experiments is ever to be made . in this context , the recent report of a spin - ss phase in a spin - half frustrated daf by chen _ et al._,@xcite who have relied on a novel tensor - product algorithm , is an important result .
however , in view of the first - order transitions observed in the related case of the shastry - sutherland model , a systematic investigation of the possibility of phase separation ( ps ) is still required . ,
is defined on a comb " geometry : holes hop ( with amplitude @xmath2 ) through the links indicated by dashed lines and repel , with strength @xmath3 , one another along the vertical nearest - neighbor links indicated by solid lines .
, scaledwidth=18.0% ] in this paper , we investigate the interplay between ss order and ps in the frustrated daf analyzed in ref . .
we begin by estimating the energetic gains behind ps and supersolidity for hard - core bosons on the square lattice by relying on strong coupling arguments , and introduce an indicator of the instability toward ps .
we then proceed to the analysis of effective bosonic models obtained from the application of the contractor renormalization ( core ) algorithm@xcite to the daf investigated in ref . .
the so - obtained effective hamiltonians are studied by performing self - consistent cluster mean - field theory ( scmft ) calculations and tendency toward ps is gauged through means of the aforementioned indicator .
in this section , we analyze the interplay between ps and supersolidity in models of hard - core bosons on the square lattice , onto which the low - energy physics of the spin model considered in the remainder of this paper can be mapped .
we start by considering the simplest model of hard - core bosons on the square lattice , the so - called @xmath4 model : @xmath5 @xmath6 is the occupation number operator for _ holes_@xcite at site @xmath7 ( a hard - core constraint is imposed ) , whose density is controlled by the chemical potential @xmath8 . here , only nearest - neighbor ( nn ) hopping ( @xmath2 ) and interaction ( @xmath3 ) terms are considered but , as will be discussed later , the effective model derived for the spin model eq . further comprises longer - ranged and/or multi - body couplings . obtained from eds on the geometry depicted in fig .
[ fig : dwall](b ) , for clusters comprising from @xmath9 ( @xmath10 ) to @xmath11 ( @xmath12 ) sites . , scaledwidth=33.0% ] the model eq . had been for many years conjectured to support ss phases ( see refs . and
references therein ) , but a more systematic numerical analysis@xcite later disproved earlier evidences in favor of this scenario and showed that , instead , ps prevents the occurrence of supersolidity .
an intuitive explanation for such behavior was put forward in ref . by relying on strong coupling arguments . following this analysis,@xcite holes [ or also particles in the case of eq . ; @xcite however , the effective models to be analyzed later lack particle - hole symmetry ] doped into the cbs ground - state of eq . for @xmath13 at half - filling would delocalize with an effective hopping amplitude proportional to @xmath14 and eventually condense , giving origin to ss order .
however , this last conclusion is flawed in that it ignores the possibility of ps .
indeed , in the strongly interacting regime domain wall [ dw , depicted as the shaded region in fig .
[ fig : dwall](a ) ] formation is energetically favored for the model eq .
: the energetic gain per _ doped _ hole ( we denote the number of _ doped _ holes by @xmath15 ) is _ linear _ in @xmath2 under these circumstances , @xmath16 with @xmath17$].@xcite since one of our primary goals in the present work is to investigate the interplay between ps and supersolidity , so to be able to decide which among the two possibilities take place for the spin model herein analyzed [ eq . ]
, we would like to obtain a more accurate estimate for @xmath18 ; in other words , we would like to pinpoint the actual values assumed by @xmath19 $ ] . in achieving this goal
we completely ignore fluctuations in cbs - ordered regions away from dws , a supposedly good approximation for @xmath13 , and consider a simplified @xmath4-like model at half - filling ( quarter - filling for _ doped _
holes ) defined on the comb " geometry depicted in fig . [ fig : dwall](b ) .
in such a simplified model hopping processes with amplitude @xmath2 only take place in between nn sites linked by the comb s teeth " and interaction @xmath3 is only active for holes sitting on nn sites along the backbone " [ see fig . [
fig : dwall](b ) ] .
] , that allow extra holes to delocalize in a cbs background by leapfrogging on the other sublattice ( c ) . adapted from ref . . ,
scaledwidth=12.5% ] in fig .
[ fig : e0_dwall ] we plot results for @xmath20 , as a function of @xmath21 , obtained from exact diagonalizations ( eds ) of the just discussed simplified model on the comb geometry depicted in fig .
[ fig : dwall](b ) , for clusters comprising up to @xmath11 sites ( thus , up to @xmath12 doped holes ) .
we first notice that @xmath22 in the limit of large @xmath23 . on the other hand ,
since fluctuations away from the dw are ignored in our analysis , we expect our ed results to _ underestimate _ @xmath24 for small values of @xmath23 .
nonetheless , we keep this limitation in mind and throughout the remainder of this paper rely on ed results in estimating the dw energy even in the weakly interacting regime @xmath25 . in doing so , we take advantage of the very small finite size effects in the data shown in fig .
[ fig : e0_dwall ] and obtain @xmath24 from eds on finite clusters . in a previous work,@xcite
we have shown that the instability toward ps is suppressed in models of hard - core bosons on the square lattice that include , in addition to the terms comprised in the @xmath4 model [ eq . ] , the correlated hopping processes with amplitudes @xmath26 and @xmath27 depicted in fig .
[ fig : leapfrog ] .
indeed , such correlated hopping terms have been shown@xcite to favor supersolidity by allowing doped holes to delocalize on top of a cbs by leapfrogging " on the other sub - lattice .
we wish now to devise a criterion for determining how large should the amplitudes @xmath26 and @xmath27 be so to inhibit dw formation and thus stabilize a ss . in doing so ,
we once more rely on strong coupling arguments , and estimate the energetic gain associated to the leapfrog processes represented in fig . [
fig : leapfrog ] as @xmath28 that is , @xmath29 is simply the ground - state energy of a single hole doped into a
frozen " cbs , an approximation expected to hold for @xmath13 , hopping via the processes with amplitude @xmath26 and @xmath27 [ fig .
[ fig : leapfrog](a - b ) ] . as it happens for our estimate @xmath30 obtained in sec .
[ sec : dw ] , we expect @xmath29 as given by eq .
to underestimate the actual energetic gain associated to the leapfrog processes . )
] , with couplings : @xmath31 ( thick vertical lines ) , @xmath32 ( thiner in - layer lines ) and @xmath33 ( dashed lines ) .
a magnetic field @xmath34 promotes singlets ( vertical pairs of filled circles ) to triplets ( pairs of open circles ) ; a cbs configuration at half - filling is depicted .
( b ) @xmath35 cluster for scmft : interactions ( thick black lines ) involving only in - cluster sites ( dark - filled circles ) are treated exactly while couplings to the environment ( grey lines ) in a mf way .
although only nn bonds are depicted , the effective model from core also includes longer - ranged terms .
, scaledwidth=15.0% ] we combine the just presented analysis and the one discussed in sec . [
sec : dw ] concerning dw formation and introduce an indicator for analyzing the interplay between supersolidity and ps : the difference between @xmath29 [ eq . ] and @xmath30 ( obtained from eds by using @xmath23 as input ; see secs .
[ sec : dw ] ) , our estimates for the energetic gains respectively associated to each of these possibilities .
since both estimates are obtained from strong coupling analysis , the indicator @xmath36 can only be expected to be accurate in the limit of @xmath13 . however , we keep this limitation in mind and in the analysis to be performed in sec . [
sec : frustrated ] , we rely on @xmath37 as an indicator even for couplings @xmath25 .
in most of the lattice models that have , so far , been shown to display ss properties the effective repulsion ( necessary to destabilize the uniform superfluid and induce a ss state ) stems from the presence of nearest - neighbor repulsive terms in xxz hamiltonians.@xcite while these anisotropic models are interesting from a theoretical perspective , their strong anisotropic character renders them unrealistic for antiferromagnetic mott insulators .
more promising in this sense is the frustrated spin - half hamiltonian analyzed in ref . ,
defined on a bilayer geometry [ fig .
[ fig : bilayer](a ) ] @xmath38 \\ + & \sum_{i }
\left [ j_{\perp } \vec{s}_{i,1 } \cdot \vec{s}_{i,2 } - h \sum_{\alpha=1,2 } s^{z}_{i , \alpha } \right]~. \label{eq : spin - dimer } \end{split}\ ] ] @xmath39 denotes nn sites in each square layer @xmath40 of the frustrated bilayer depicted in fig .
[ fig : bilayer](a ) .
@xmath31 couples spins in different layers to build the basic dimers of the model ( we set @xmath41 ) .
the applied magnetic field @xmath34 acts as a chemical potential , promoting spin - dimers from a singlet ( hole ) to a triplet ( triplon ) state .
effective interactions appear as the result of in - layer @xmath32 and frustrating @xmath33 antiferromagnetic couplings .
we remark that the lattice depicted in fig .
[ fig : bilayer](a ) remains invariant if every other spin - dimer is rotated by @xmath42 and thus eq .
( [ eq : spin - dimer ] ) is invariant under the transformation @xmath43 , with the consequence that the phase diagram is symmetric about the line @xmath44 . in studying the model of eq .
( [ eq : spin - dimer ] ) , we adopt an approach similar to the one employed in our previous work ref . , where a related unfrustrated model was investigated and to which the reader is referred for details,@xcite and derive an effective bosonic model by relying on the core algorithm.@xcite we consider spin - dimers connected by @xmath31 as elementary blocks and select the singlet @xmath45 $ ] and the @xmath46 triplet @xmath47 as the block states in the core expansion : for all parameters in eq .
( [ eq : spin - dimer ] ) considered in the present work , @xmath48 $ ] , this choice is justified by the large reduced density - matrix weights associated to such block states and by the rapid convergence of effective couplings for increasing range in the expansion.@xcite effective couplings are derived by diagonalizing clusters of coupled dimers and by projecting a matching number of low - lying cluster eigenstates onto the basis formed by tensor products of the retained block states , @xmath49 and @xmath50.@xcite the effective bosonic hamiltonian thus obtained is essentially identical to the one derived for the anisotropic model studied in ref .
, _ only _ the magnitudes for each coupling being different . similarly , the effective model obtained here is _ not _ invariant under particle - hole transformation and , in particular , amplitudes for leapfrog processes " are non - zero only when _
holes _ are involved . from this last observation
we expect that only hole - doped ss phases can be stabilized in the spin - dimer model eq .
( [ eq : spin - dimer ] ) and conclude that the effective hamiltonian is more conveniently expressed in terms of hole operators ( @xmath51 is the occupation number for holes at the dimer - lattice site @xmath52 ) .
we therefore adopt the same notation as in our previous work,@xcite to which the reader is referred for a complete list of single- and multi - hole interactions and hopping processes [ see eqs .
( 8 , b1-b5 ) in ref . ] .
we also remark that the effective hamiltonian preserves the symmetry of the original model eq .
( [ eq : spin - dimer ] ) and remains invariant under @xmath53 .
effective hamiltonians resulting from core are often complex and different strategies may be pursued in trying to extract physically sound results from them .
one possibility is the mean - field ( mf ) theory of ref .
, which reproduces semi - quantitatively the results of quantum monte carlo ( qmc ) simulations@xcite for the anisotropic spin - dimer model considered therein .
however , mf calculations are known to overestimate the extent of ss phases and it would be desirable to include , at least partially , effects due to quantum fluctuations . from this perspective , the scmft@xcite that partially takes local quantum fluctuations into account and has been recently applied to the @xmath4 model for hard - core bosons on the triangular lattice@xcite seems particularly well suited for our purposes .
indeed , the extent of the ss phase in the ground - state phase diagram obtained by applying scmft to the @xmath4 model on the triangular lattice@xcite compares considerably better with results from qmc simulations@xcite than what is found from a more conventional mf approach.@xcite [ squares , eq .
( [ eq : rho0 ] ) ] , cbs structure factor @xmath54 [ circles , eq .
( [ eq : cbs ] ) ] and magnetization density [ triangles , eq . ( [ eq : mz ] ) ] for the effective core hamiltonian for eq .
( [ eq : spin - dimer ] ) with couplings @xmath55 considered in ref .
: ( a ) @xmath56 and ( b ) @xmath57 .
successive phases for increasing magnetic field @xmath34 are labeled as : spin - gapped ( @xmath58 ) , condensate ( bec ) , supersolid ( ss ) , checkerboard solid ( cbs ) and fully polarized ( @xmath59 ) .
, scaledwidth=22.0% ] [ squares , eq .
( [ eq : rho0 ] ) ] , cbs structure factor @xmath54 [ circles , eq .
( [ eq : cbs ] ) ] and magnetization density [ triangles , eq . ( [ eq : mz ] ) ] for the effective core hamiltonian for eq .
( [ eq : spin - dimer ] ) with couplings @xmath55 considered in ref .
: ( a ) @xmath56 and ( b ) @xmath57 .
successive phases for increasing magnetic field @xmath34 are labeled as : spin - gapped ( @xmath58 ) , condensate ( bec ) , supersolid ( ss ) , checkerboard solid ( cbs ) and fully polarized ( @xmath59 ) .
, scaledwidth=22.0% ] .
( a ) extent of cbs ( @xmath60 ) and ss ( @xmath61 ) phases [ maximum minus minimum value of the field @xmath34 leading to the corresponding phase for given parameters @xmath62 .
( b ) value of the structure factor [ eq . ( [ eq : cbs ] ) ] at the cbs plateau . in ( a ) and ( b ) , symbols indicate results by chen _
et al._@xcite and lines the here obtained results .
, scaledwidth=30.0% ] scmft is applied by diagonalizing the effective core hamiltonian on the @xmath63 cluster depicted in fig .
[ fig : bilayer](b ) . in setting the cluster s hamiltonian ,
in - cluster interactions are treated _ exactly _ while couplings to the environment in a self - consistent way : for instance , a given interaction connecting sites @xmath52 and @xmath64 contributes a term proportional to @xmath65 for each in - cluster bond [ thick black lines in fig . [
fig : bilayer](b ) ] and with mean - field terms of the form @xmath66 $ ] for bonds " connecting the cluster to its environment [ grey lines in fig . [ fig : bilayer](b ) ] . at each step , the ground - state for the cluster hamiltonian is calculated and expectation values @xmath67 , @xmath68 at every site @xmath52 computed ; these are then used in setting the mean fields for the next iteration , until convergence is achieved ( see ref . for details ) . in this way
, we compute the condensate density at the point @xmath69 @xmath70 the cbs structure factor ( normalized per site ) @xmath71 and the magnetization along the field direction @xmath72 in fig . [ fig : mf ] we plot these quantities as a function of the magnetic field @xmath34 for couplings considered in ref . , @xmath73 and @xmath57 .
we first notice that the overall agreement between our results and the data presented in ref .
is remarkably good.@xcite for the least frustrated case of @xmath73 [ fig .
[ fig : mf](a ) ] , the system first undergoes a quantum transition from a spin - gapped ( equivalent to a trivial bosonic mott insulator with zero - filling for triplons , @xmath74 in fig .
[ fig : mf ] ) to a bec phase at the lower critical field @xmath75 , and then from the bec to a fully polarized phase ( trivial mott insulator with unitary triplon filling , @xmath76 ) at the upper critical field @xmath77 . more interestingly
, additional cbs and ss phases are stabilized for the more frustrated case of @xmath78 [ fig .
[ fig : mf](b ) ] .
the existence of a ss phase at the low - field boundary of the cbs plateau , with finite values for both @xmath79 and @xmath80 is at least partially due to the presence of correlated hoppings for _ holes _ in the effective core hamiltonian .
indeed , no ss phase is observed for an effective model " obtained by setting @xmath81 [ fig .
[ fig : leapfrog](a - b ) ] while keeping all the other effective couplings unchanged .
this situation is to be contrasted with the first - order transition from cbs to bec at higher fields , explained by the _ vanishing _
amplitudes for correlated hoppings for _ triplons _ for all values @xmath48 $ ] .
we proceed by varying the frustrating coupling @xmath33 while fixing @xmath82 . in fig .
[ fig : extent](a ) we plot the extent of the ss , cbs phase , respectively @xmath83 [ @xmath84 ( @xmath85 ) denotes the upper ( lower ) boundary of the ss , cbs phase ] as a function of @xmath86 . in order to further gauge the accuracy of the here employed core - scmft approach in fig .
[ fig : extent](a ) our results for @xmath61 and @xmath60 are compared against those from ref . and in fig .
[ fig : extent](b ) we plot both our results and those from ref .
for the structure factor @xmath54 at the cbs plateau .
excellent agreement is found in both cases and we further remark that our results for @xmath54 in the cbs phase in fig .
[ fig : extent](b ) confirm that quantum fluctuations are indeed partially taken into account by scmft : in contrast to what happens with the semi - classical mf approach employed in ref .
, here the value of @xmath54 at the plateau is somewhat reduced from its classical value @xmath87 . at this point , and despite of its aforementioned attractive features , it is important to have in mind an important limitation of the here employed scmft procedure : since calculations rely on diagonalizations of a @xmath88 cluster [ fig .
[ fig : bilayer](b ) ] , only _ homogeneous _ solutions , displaying order consistent with at most quadrupling of the unit cell , are obtainable .
this excludes _ inhomogeneous _ solutions such as those associated with the presence of dws [ fig .
[ fig : dwall](a ) ] and has the consequence that our combined core - scmft approach is insensitive to the occurrence of ps . in
what follows , we rely on the strong coupling analysis presented in sec .
[ sec : ps ] and analyze the interplay between ps and spin - ss order in the phase diagram of the model eq .
( [ eq : spin - dimer ] ) .
we now turn our attention to the obtention of a global @xmath32 @xmath33 phase diagram that may guide the experimental search for realizations of spin - supersolidity and therefore extend our analysis by varying @xmath32 in eq .
( [ eq : spin - dimer ] ) . in fig .
[ fig : range](a ) we plot @xmath61 as a function of @xmath48 $ ] , obtained from the combined core - scmft procedure .
these results suggest that , far from being a rare occurrence , spin - supersolidity is widespread throughout the parameter space and can extend over fairly wider ranges of @xmath34 than it is observed for the value @xmath82 [ couplings considered in ref .
are highlighted in fig .
[ fig : range](a ) ] . however , under the light of our discussion in sec .
[ sec : ps ] concerning ps in systems of hard - core bosons on a lattice , some caution is required in drawing conclusions from the results shown in fig .
[ fig : range](a ) . following the discussion in sec .
[ sec : ps ] , we evaluate @xmath89 , our indicator for analyzing the interplay between ps and supersolidity for hard - core bosons on the square lattice , as a function of @xmath48 $ ] for eq .
( [ eq : spin - dimer ] ) .
@xmath29 , our strong coupling estimate for the energetic gain associated to occurrence of ss order , is readily obtained by plugging the amplitudes for the leapfrog processes @xmath26 and @xmath27 [ fig .
[ fig : leapfrog](a - b ) ] obtained from core into eq . .
on the other hand , the estimate @xmath90 for the energy associated to ps is obtained from numerical eds for the simplified model for dws defined on the comb geometry " discussed in sec .
[ sec : dw ] by using the effective ratio @xmath23 , as obtained from the core expansion for each set @xmath91 in eq .
( [ eq : spin - dimer ] ) , as an input .
eds are performed on a small cluster comprising @xmath92 sites ( @xmath93 doped holes ) : as mentioned in sec .
[ sec : dw ] , this is justified by the absence of sizable finite size effects for the data displayed in fig .
[ fig : e0_dwall ] .
( see main text ) for the frustrated daf eq .
( [ eq : spin - dimer ] ) .
the symmetry @xmath43 has been explored in obtaining the data .
regions where supersolidity [ ps ] is expected , where @xmath94 [ @xmath95 are marked by the label ss [ ps ] .
dashed lines indicate threshold values for a cbs / ss to appear at the mean - field level .
( b ) @xmath96 , as obtained from eds on an @xmath92 ( @xmath93 doped holes ) site cluster with the comb geometry depicted in fig .
[ fig : dwall](b ) , for the model eq .
( [ eq : spin - dimer ] ) .
contour levels for @xmath23 ( obtained from the core expansion ) are indicated by thin lines and the values @xmath97 , @xmath98 , @xmath99 and @xmath100 are highlighted . in both panels ,
circles indicate couplings investigated by chen _
et al._@xcite and the thick line couplings yielding the threshold value @xmath101 . , title="fig:",scaledwidth=45.0% ] ( see main text ) for the frustrated daf eq .
( [ eq : spin - dimer ] ) .
the symmetry @xmath43 has been explored in obtaining the data .
regions where supersolidity [ ps ] is expected , where @xmath94 [ @xmath95 are marked by the label ss [ ps ] .
dashed lines indicate threshold values for a cbs / ss to appear at the mean - field level .
( b ) @xmath96 , as obtained from eds on an @xmath92 ( @xmath93 doped holes ) site cluster with the comb geometry depicted in fig .
[ fig : dwall](b ) , for the model eq .
( [ eq : spin - dimer ] ) .
contour levels for @xmath23 ( obtained from the core expansion ) are indicated by thin lines and the values @xmath97 , @xmath98 , @xmath99 and @xmath100 are highlighted . in both panels ,
circles indicate couplings investigated by chen _
et al._@xcite and the thick line couplings yielding the threshold value @xmath101 .
, title="fig:",scaledwidth=45.0% ] we plot @xmath89 as a function of @xmath48 $ ] in fig .
[ fig : range](b ) and assume that two conditions must be simultaneously fulfilled for ss phases to exist for the model eq .
( [ eq : spin - dimer ] ) : ( i ) a ss must be observed within core - scmft _ and _ ( ii ) @xmath94 .
values of @xmath91 leading to @xmath101 , the threshold value for a ss phase to appear , are indicated by the thick continuous curve in fig .
[ fig : range ] .
we notice that not all values of @xmath102 yielding a spin - ss phase within our core - scmft approach fulfill @xmath94 and expect ps to take place under these circumstances instead . despite of the fact that the condition @xmath94 considerably shrinks the size of the region expected to support ss phases from a pure core - scmft analysis ,
supersolidity is still observed for a wide range of couplings in eq .
( [ eq : spin - dimer ] ) [ fig .
[ fig : range](a ) ] , possibly realizable in real magnets .
intriguingly , we notice that the parameters ( circles in fig .
[ fig : range ] ) for which a spin - ss phase has been detected by chen _
et al_.,@xcite and also by the pure core - scmft analysis devised here ( figs .
[ fig : mf ] and [ fig : extent ] ) , fail to satisfy @xmath94 [ fig .
[ fig : range](b ) ] . although we can not exclude the possibility that our criterion , that rigorously applies only in the limit @xmath13 , is too stringent for the frustrated model eq .
( [ eq : spin - dimer ] ) , we remark that a ss phase is obtained within our core - scmft approach@xcite and in ref . even for couplings
@xmath103 , where our strong coupling arguments become accurate [ the ratio @xmath104 diverges toward the line @xmath105 ; see fig . [
fig : range](b ) , where contour levels for @xmath106 are plotted as thin continuous lines ] .
this inconsistency suggests that both core - scmft and the novel algorithm employed in ref .
are insensitive to the instability toward ps in systems of hard - core bosons on the square lattice and that the obtention of ss phase for @xmath107 is _
spurious_.@xcite it would therefore be important to further test the ability of the algorithm employed in ref . to detect ps in bosonic lattice models by , for instance , checking how it compares to qmc for the unfrustrated model studied in refs . regarding this issue .
summarizing , we have studied a spin - half frustrated bilayer model by combining core and scmft .
our results reveal the presence of a spin - ss phase under applied magnetic field , which appears at the edge of a half - saturated magnetization plateau and is stabilized by a
leapfrog mechanism".@xcite we address the interplay between supersolidity and instability toward ps , that precludes the emergence of spin - ss phases , by devising a quantitative criterion based on strong coupling arguments .
this criterion is generically applicable to systems of hard - core bosons on the square lattice , and it would be interesting to further assess its validity by investigating models where the interplay between ps and ss can be independently analyzed . by relying on this criterion , we obtain a global phase diagram for the frustrated spin - dimer antiferromagnet considered herein , and show that a spin - ss phase is stable against ps for couplings realizable in real magnets .
we expect that our results may guide the experimental search for systems exhibiting spin - supersolidity .
we acknowledge fruitful discussions with c. d. batista and m. troyer , as well as funding from the french anr program anr-08-jcjc-0056 - 01 , from arc ( australia ) , from the snf and from manep ( switzerland ) .
nl acknowledges lpt toulouse for hospitality . | we investigate the competition between spin - supersolidity and phase separation in a frustrated spin - half model of weakly coupled dimers .
we start by considering systems of hard - core bosons on the square lattice , onto which the low - energy physics of the herein investigated spin model can be mapped , and devise a criterion for gauging the interplay between supersolid order and domain wall formation based on strong coupling arguments .
effective bosonic models for the spin model are derived via the contractor renormalization ( core ) algorithm and we propose to combine a self - consistent cluster mean - field solution with our criterion for the occurrence of phase separation to derive the phase diagram as a function of frustration and magnetic field . in the limit of strong frustration ,
the model is shown to be unstable toward phase separation , in contradiction with recently published results . however
, a region of stable supersolidity is identified for intermediate frustration , in a parameter range not investigated so far and of possible experimental relevance . |
recent observations of the cosmic microwave background radiation ( cmbr ) have found statistically significant anisotropy at 05 angular scales ( see @xcite , and references therein ) . in earlier letters ( @xcite , hereafter paper i , and @xcite , hereafter paper ii ) , we reported on observations of cmbr anisotropy that included two particularly bright features .
the presence of bright , unresolved sources suggested the possibility of foreground point - source contamination , or of non - gaussian fluctuations in the cmbr . in this letter , we present our analysis of these features and a discussion of other work studying their properties .
the msam1 instrument has been described in detail elsewhere ( @xcite ) ; only the most relevant features are given here .
msam1 is a balloon - borne off - axis cassegrain telescope with a nutating secondary mirror and a four - band bolometric radiometer .
the detectors have center frequencies of 5.7 , 9.3 , 16.5 , and 22.6 each with @xmath1 bandwidth .
the beam size of the telescope is 28 fwhm , roughly independent of frequency .
the secondary executes a horizontal four - position square - wave chop ( i.e. , center , left , center , right ) with an amplitude of @xmath2 on the sky .
the telescope is pointed using a combination of torque motors and reaction wheels , referenced to a gyroscope .
the absolute sky location is determined by star camera observation .
the data discussed here were collected during two flights of the msam instrument from palestine , texas ; the first on 1992 june 5 ( msam1 - 92 ) , the second on 1994 june 2 ( msam1 - 94 ) .
these observations have been described in papers i and ii , and are only summarized here . in each flight , the antenna patterns for all four bands were measured by rastering over jupiter .
we determine the location of the antenna pattern within the star camera s field - of - view to within 2 ( post - flight ) by simultaneously observing jupiter in the star camera and the main telescope .
gyroscope drift increases our absolute pointing uncertainty to @xmath3 .
we calibrate our spectral flux density measurements using the measured surface temperatures ( @xcite ) together with jupiter s apparent diameter at the epoch of flight ( @xcite , @xcite ) .
the calibration has a 10% overall uncertainty , with relative errors of 14% between bands .
the cmbr observations were made at roughly constant declination @xmath4 , with the central beam covering right ascensions 144169 and 172203 during the msam1 - 92 flight .
this corresponds to @xmath5 square degrees of sky coverage .
the goal of the msam1 - 94 flight was to observe the identical field as msam1 - 92 ; this was nearly achieved , with an average declination difference of @xmath6 .
the coverage in right ascension was somewhat less , extending from 150170 , and 173200 .
except for [ sec : counterparts ] , all results reported in this letter are based on the msam1 - 92 observations .
the initial data reduction used in this analysis has been reported in papers i and ii ; only the salient features are reviewed here .
note that this letter is based on the reanalyzed version of the msam1 - 92 data including the full instrument covariance matrix ( @xcite ) .
two nearly statistically independent demodulations are formed from the data .
the single - difference demodulation subtracts the right beam from the left beam , and ignores the central beam data ; the double - difference demodulation subtracts the average of the left@xmath7right beams from the central beam .
the single - difference demodulation has an antisymmetric two - lobe response pattern , while the double - difference demodulation has a symmetric three - lobe pattern .
the data are binned by position and angular orientation of the demodulated antenna pattern on the sky . to remove instrumental drift ,
a slowly varying function of time is fit simultaneously with the bins .
the bin size is 012 in position , and 10 in angular orientation .
the dominant sources of brightness contrast expected in our bands are anisotropy in the cmbr , and thermal emission from warm ( @xmath8k ) interstellar dust ( @xcite , @xcite ) . before searching for compact features
, we use one of our spectral degrees of freedom to make a simple correction for the presence of dust emission . unlike our previous analyses for cmbr anisotropy
, here we do not impose a full spectral model on the data .
rather , after a minimal correction for dust contamination , we model the spatial structure of several bright , compact features , and study the spectral nature of those features .
we estimate and subtract the contribution of foreground interstellar dust emission using the 22.6 band , assuming a @xmath9 emissivity rayleigh - jeans spectrum ( the noise estimates in the remaining three bands are adjusted to reflect this subtraction ) . in making this correction
, we assume that the 22.6 band is an uncontaminated monitor of dust .
varying the exponent of the power law emissivity of the dust by @xmath10 has a negligible effect on the final results , compared with the statistical uncertainties .
the dust correction is most important for the 16.5 band , and has little effect at the lower frequencies . after removing the estimated dust contamination ,
we search the remaining 3 frequency bands for unresolved features using an @xmath11statistic to measure significance .
each band is searched independently , and then the source lists are compared . since the telescope is scanned at roughly constant declination , we parameterize the location of the features as a function only of right ascension , and constrain the central declination of the feature to follow a slow linear function of right ascension .
we begin by calculating the raw @xmath12 of the data in each band , @xmath13 where @xmath14 are the binned data with associated noise covariance @xmath15 , and the sum is over both demodulations as well as bins .
@xmath16 is one of four arbitrary offset parameters corresponding to the two demodulations and two portions of the flight , appropriate to datum @xmath17 .
then , as a function of right ascension @xmath18 , we find the flux density @xmath19 that minimizes @xmath20 \wgt \left [ \rawj - o_j - s_\nu(\alpha ) \beamj \right],\ ] ] where @xmath21 is the value of the beam map ( single or double difference , as appropriate ) centered on observation point @xmath17 , evaluated at right ascension @xmath18 .
the significance of the fitted flux density is determined by forming the statistic , @xmath22 where @xmath23 is the number of degrees of freedom remaining in the data after fitting out the source .
this statistic should be distributed according to an @xmath11-distribution with 1 and @xmath23 degrees of freedom ( @xcite ) , which allows us to calculate the cumulative probability @xmath24 of drawing a value greater than or equal to @xmath11 if the null hypothesis that the actual value @xmath25were true .
values of @xmath26 are highly unlikely under the null hypothesis , and correspond to significant detections of compact features in the data .
figure [ fig : significance ] shows @xmath27 and @xmath24 for each of the three frequency bands .
+ since we are plotting the cumulative probability of the null hypothesis being true , lower values of @xmath24 correspond to more significant detections .
we have multiple lobe beam patterns , so a bright unresolved feature will alias to several values of right ascension separated by the 40beam throw .
thus , the clusters of peaks near 15 and 193 do not correspond to multiple features , but rather the aliasing of a single feature into the side beams .
we build a list of candidate unresolved features for each band by iteratively searching for and subtracting the most significant feature , defined here as the maximum value of @xmath27 . after each feature
is subtracted , @xmath19 , and @xmath27 are re - evaluated to help suppress the aliasing in right ascension .
we continue this procedure until @xmath28 , corresponding to @xmath29 .
three candidate features were identified this way for the 5.7 and 9.3 bands .
we allow each feature to be positive or negative , but the three brightest features are all positive .
no features were found at 16.5 which satisfy the @xmath30 requirement .
table [ table : clean ] lists the results of this search , along with the corresponding central declinations .
rrrrr 1497 & @xmath78197 & 62 & @xmath31 & @xmath32 1929 & @xmath78187 & 25 & @xmath33 & @xmath34 1992 & @xmath78185 & 14 & @xmath35 & @xmath36 1496 & @xmath78197 & 27 & @xmath37 & @xmath38 1930 & @xmath78187 & 12 & @xmath39 & @xmath40 1994 & @xmath78185 & 12 & @xmath41 & @xmath42 to determine the spatial extent of each feature , we fit the data to a gaussian shaped surface brightness model , @xmath43 , with the central brightness @xmath44 , right ascension @xmath18 , and width @xmath45 as free parameters .
the fits are performed using the levenberg - marquardt method ( @xcite ) .
the total flux density of the feature is related to the central brightness by @xmath46 .
the fits near 15 indicate a clearly resolved feature in both the 5.7 and 9.3 bands , with @xmath47 .
the results of these fits are given in table [ table : fit1 ] .
there are substantial correlations between and @xmath44 , so these are also listed . from @xmath44 , we can also infer an equivalent peak @xmath48k ( 5.7 ) , @xmath49k ( 9.3 ) , near 15 .
the fits near 193 and 199 do not show evidence for significant spatial extent .
we can infer 68% ( 1@xmath50 ) upper bounds on from the estimated parameter variance of the fit ; these give @xmath51 at 193 and @xmath52 at 199 , consistent in both bands .
because of the close proximity of the 193 and 199 features , there is some correlation between their measured flux densities . to account for this ,
the fits are repeated , omitting while simultaneously fitting for @xmath18 and @xmath53 of both features .
the results of these fits are shown in table [ table : fit2 ] .
rlrr @xmath54 & 1496@xmath55002 & & : & 36@xmath558 & @xmath56 & @xmath57 : & 158@xmath5524 kjy / sr & @xmath58 & @xmath59 @xmath18 : & 1495@xmath55003 & & : & 33@xmath5513 & @xmath60 & @xmath61 : & 137@xmath5540 kjy / sr & @xmath62 & @xmath63 @xmath64 : & @xmath6577 kjy / sr & rlrrr @xmath66 : & 1929@xmath55002 & & & @xmath67 : & 1992@xmath55003 & @xmath68 & & @xmath69 : & [email protected] jy & @xmath70 & @xmath71 & @xmath72 : & [email protected] jy & @xmath73 & @xmath74 & @xmath75 @xmath66 : & 1932@xmath55003 & & & @xmath67 : & 1993@xmath55003 & @xmath76 & & @xmath77 : & [email protected] jy & @xmath78 & @xmath79 & @xmath80 : & [email protected] jy & @xmath81 & @xmath70 & @xmath82 @xmath83 : & [email protected] jy & & & @xmath84 : & [email protected] jy & & & @xmath78 while there are no bright features detected at 16.5 , we can find limits for the corresponding features found at the lower frequencies .
for the first feature , we fit using our extended model , fixing @xmath85 and @xmath86 , while for the other features , we fit assuming completely unresolved sources with @xmath87 and @xmath88 fixed . in all cases ,
the results are consistent with zero flux , and upper bounds can be inferred from the variance . these are also listed in tables [ table : fit1 ] and [ table : fit2 ]
we now compare the measured spectra of the features with four simple models for astrophysical sources : cmbr anisotropy , free - free emission with @xmath89 , synchrotron emission with @xmath90 , and cold dust emission with @xmath91 .
the last of these is motivated by the possible detection of a cold dust contribution to the galactic spectrum observed by the /firas experiment ( @xcite , @xcite ) .
each of the models is fit to the values listed in tables [ table : fit1 ] and [ table : fit2 ] , the results of which are displayed in figure [ fig : spectra ] .
the @xmath12 for the spectral fits are in table [ table : spectra ] .
the relative calibration uncertainty is small compared to the statistical uncertainties , and is not included in the @xmath12 values .
+ ccccc 150 & 0.8 & 4.5 & 2.0 & 40 .
193 & 0.3 & 0.7 & 2.0 & 20 .
199 & 2.4 & 1.7 & 4.3 & 6.9 all three features are in agreement with the cmbr model .
the first two are significantly discrepant with a cold dust spectrum , while the feature near 199 is marginally consistent with that spectrum .
furthermore , this location corresponds to the brightest warm dust emission in our field . based on this
, we can not rule out the possibility that the 199 feature is be due to galactic dust emission . from these data alone , either free - free or synchrotron emission gives an acceptable fit , and could explain the detections . this would , however , require implausibly large amplitudes for emission at lower frequencies .
for example , the best fit free - free spectra , extrapolated to 40 ghz ( q - band ) , imply antenna temperatures of 3.4 , 1.1 , and 0.9 mk for the three features ; measurements made at saskatoon of the same field strongly rule out such bright signals ( @xcite ) .
the limits on synchrotron emission are even stronger , since the extrapolated temperatures are roughly 3 times greater .
recent measurements of h@xmath18 emission in this part of the sky imply the free - free contribution to the msam1 signal at 5.7 is @xmath92500 jy / sr rms ( @xcite ) , several orders of magnitude smaller than the signal we report .
we consider the possibility that these detections correspond to compact radio ( or infrared ) point sources .
note that the power law models corresponding to free - free and synchrotron also can be interpreted as flat - spectrum or steep - spectrum radio spectra , respectively , typical of extragalactic radio sources .
no counterparts for the features were found within one beamwidth of the feature locations in the 5 ghz s5 polar cap survey ( @xcite ) .
for all three features , the best - fit flat spectra predict 5 ghz fluxes greater than 3 jy , while the best - fit steep spectra predict fluxes greater than 40 jy .
the s5 survey is estimated to be complete at 5 ghz down to 250 mjy , providing a strong limit on any flat- or steep - spectrum radio counterpart to the detections .
a similar search was performed with the 100 faint source catalog ( @xcite ) .
a total of 5 objects ranging in flux from 0.92.6 jy were found within one beamwidth of any one of the three features , consistent with the average fsc source density of @xmath93sq .
the contribution of these sources to the msam signal is shown to be negligible using the 1.25 mm to 100 flux ratio , @xmath94 , determined from a complete sample of galaxies , together with a relatively flat @xmath95 spectrum locally around 1.25 mm ( @xcite ) .
the greatest contribution is at 16.5 , where it is estimated to be @xmath96 jy for each of the latter two features , and @xmath97 mjy for the 15 feature .
the contributions at 9.3 are a factor of 5 smaller than these , while at 5.7 they are a factor of 30 smaller .
these estimates are reinforced with direct observations at 90 ghz of known iras sources in the vicinity of the 15 feature , none of which show significant flux density above the rms noise of 2550 mjy ( @xcite ) . in 1994
, we flew the msam instrument a second time to observe the same field as the 1992 flight .
while the sky overlap between msam1 - 94 and msam1 - 92 was not perfect , it was sufficient to permit an overall comparison of the measurements ( @xcite ) .
we have repeated the fits of [ sec : fits ] on the msam1 - 94 data to see if the bright features are present .
near 15 , there is a corresponding hot spot at @xmath98 with @xmath99 , @xmath100 kjy / sr , and @xmath101 kjy / sr ( statistical errors only ) . near 193 and 199
, there are corresponding features that are fit simultaneously as before . at 5.7 ,
the resulting parameters are @xmath102 , @xmath103 jy , and @xmath104 , @xmath105 jy . at 9.3 ,
the fit gives @xmath106 , @xmath107 jy , and @xmath104 , @xmath108 jy . the correlations in the msam1 - 94 fits are larger than the msam1 - 92 fits ; in particular , the 193 and 199 flux densities have variance correlations of @xmath109 at 5.7 , and @xmath110 at 9.3 .
the somewhat smaller flux densities in msam1 - 94 are consistent with the pointing differences between the two flights .
since the initial report of these detections in paper i , a number of groups have searched one or more of these regions for corresponding sources at different wavelengths and/or resolutions .
of particular note is the search reported by @xcite using the caltech submillimeter observatory on mauna kea .
this search , which made a deep map of the sky around the 15 feature at 4.7 using a 17 beam , places an upper limit on any point source in that field of @xmath111 jy .
another observation of this field has recently been reported by @xcite .
these observations were made from the ground at saskatoon , saskatchewan , canada , in q - band ( 3646 ghz , 1.21.5 ) at 05 resolution , a portion of which were designed to reproduce , as closely as possible , the msam observations .
a three - lobe beammap , similar to the msam double - difference demodulation , was synthesized ( the single - difference demodulation is not usable from the ground due to atmospheric noise ) .
although having somewhat less statistical weight than the msam observations , the saskatoon results correlate well with msam , including the bright feature near 15 right ascension .
the presence of bright , compact features in the msam1 - 92 data was unexpected . in paper
i , we questioned whether cmbr fluctuations obeying gaussian statistics could produce such features . to address this issue ,
we have performed a detailed monte carlo simulation of the msam observations .
the approach used is similar to that outlined in 4.3 of paper i. ] .
we generate numerical realizations of the observations , including contributions from instrument noise and cmbr temperature anisotropies , assuming the anisotropy is a gaussian random field described by a two - point correlation function , @xmath112 . in paper
i , we set limits on the overall amplitude of @xmath113 by using the likelihood ratio statistic @xmath114 ; here , we use those upper and lower bounds to generate realizations which are then searched according to the procedure of [ sec : search ] for unresolved features . unlike paper
i , here we simulate both demodulations simultaneously , using @xmath115 .
we thus determine the distribution of @xmath116 , the number of unresolved features with @xmath117 , corresponding to the range of @xmath118 fluctuations found in the data .
our measured number of sources is @xmath119 or 3 , depending on the assignment of the 199 feature to cmbr or galactic dust . for @xmath120 ,
we find the cumulative probability @xmath121 , @xmath122 , while for @xmath123 , @xmath124 , @xmath125 . from this
, we conclude that bright , compact features such as those observed are in fact consistent with our assumption of gaussian statistics .
we have also studied the distribution of @xmath116 using a more physically motivated correlation function , calculated for `` standard '' flat cold dark matter ( cdm ) with @xmath126 , and no reionization ( @xcite ) . using this model , normalized to the /dmr detection @xmath127k ( @xcite )
, we find reasonable probabilities for the observed @xmath116 : @xmath128 , @xmath129 .
thus , the detections are consistent with adiabatic fluctuation models such as cdm .
similar results have also been reported for a more generic study of 05 resolution cmbr maps ( @xcite ) .
all three features are best described as temperature anisotropies in the cmbr , although the possibility remains that one of these is due to galactic foreground dust emission . while we were initially surprised to find such bright , compact features in the data , we now understand that these are consistent with our assumption of gaussian fluctuations .
although our catalog searches were all null , we find the most compelling argument that the features are not true point - sources comes from the direct search by @xcite .
finally , the morphological agreement reported by the saskatoon experiment , consistent with a cmbr anisotropy spectrum spanning three octaves in frequency from q - band ( 1.2 ) to the msam 9.3 band , leaves little room for alternate explanations of the detection .
given the agreement with a cmbr temperature anisotropy spectrum , together with the implausability of other explanations , we conclude that at least the first two , and possibly all three of these detections , are cosmological in origin .
we would like to thank the staff of the national scientific balloon facility ( nsbf ) for their outstanding support during the two flight campaigns .
we would also like to thank e. magnier , r. rutledge , c. bennett , r. bond , j. silk , and c. lawrence for many helpful discussions of these features over the years .
this research has made use of nasa s astrophysics data system abstract and article services and ipac s xcatscan catalog scanning system , as well as software tools from the free software foundation .
the national aeronautics and space administration supports this research through grants nagw 1841 , rtop 188 - 44 , ngt 50908 , and ngt 50720 . | we report the characterization of bright , compact features in the cosmic microwave background radiation ( cmbr ) detected during the june 1992 and june 1994 balloon flights of the medium scale anisotropy measurement ( msam1 - 92 and msam1 - 94 , respectively ) .
spectral flux densities are determined for each feature at 5.7 , 9.3 , and 16.5 .
no viable counterparts for these features were found in source catalogs at 5 ghz or at 100 .
the measured spectrum of each feature is consistent with a temperature fluctuation in the cmbr .
the existence of these features is consistent with adiabatic fluctuation models of anisotropy in the cmbr . |
superconductivity with spin - triplet pairing symmetry which has been observed in a series of experiments in sr@xmath2ruo@xmath3@xcite , upt@xmath4 and some other heavy fermion complexes@xcite attracts a lot of interests in the condensed matter physics .
the spin triplet @xmath0wave pairing symmetry has been considered for unconventional superconductivity in upt@xmath5 heavy - fermion @xcite .
superconductivity has been observed in upt@xmath5 compound in two important states , @xmath6phase and @xmath7phase .
although different phases correspond to the different symmetries of the superconducting gap on the fermi surface , all of them belong to spin triplet .
the @xmath0wave order parameter has a more complex dependence on azimuthal angles in comparison with the well - known @xmath8wave order parameter for superfluid phases of @xmath9he .
the fermi surface of upt@xmath5 is highly complex with five or six complicated shape sheets .
while the pairing mechanism of upt@xmath4 is unknown , a large number of experimental and theoretical works have been done on its thermodynamic and transport property .
the josephson junction between triplet superconductors is an accurate method to study physics in triplet superconductors . in superconductor - normal metal - superconductor ( sns ) junction ,
the josephson current flows owing to the proximity effect .
recently , proximity effect in unconventional superconductor and normal metal junction has attracted much attention@xcite . the proximity effect in junctions between triplet superconductors and normal metals in the diffusive regime
have been investigated theoretically @xcite .
it is found that the proximity effect is enhanced by the mid - gap andreev bound state formed at the interface between triplet superconductor and the normal metal due to the sign change of the pair potential .
@xcite as a result , the density of states in the normal metal has a zero energy peak rather than gap . to investigate the proximity effect ,
the quasiclassical green s function method is quite useful . in this paper , we apply the quasiclassical eilenberger equation@xcite to calculate charge- and spin - currents and density of states in a triplet sns junction .
the sns josephson effect is an interesting issue because of the industrial applications such as superconducting quantum interference devices ( squid ) @xcite .
dc josephson junction has a persistent current by gradient of the macroscopic phase of superconductivity @xcite .
while the charge current is more known property in josephson junction , spin current is more interesting because of applications in spintronics .
the effect of phase difference on the spin current in the spin - triplet sis josephson junction has been studied recently @xcite .
authors of the paper @xcite obtained a polarized dissipation - less supercurrent of spins .
it is shown that the current - phase dependencies are quite different from those of the junction between conventional @xmath10-wave superconductors @xcite and high temperature @xmath11-wave superconductors @xcite . in the present paper , we investigate spin and charge josephson currents , and the density of states in junction between misorientated crystal of triplet @xmath0wave superconductors which sandwich the mesoscopic normal metal layer .
triplet @xmath12-wave pairing symmetry is considered , for example , for superconductivity in upt@xmath5 crystals @xcite .
two different geometries are corresponding to the different orientations of the crystals in the right and left sides of the interface . in geometry ( i )
the @xmath13plane in the right side has been rotated around the @xmath14axis by @xmath15 . in geometry ( ii ) , the @xmath16plane in the right side has been rotated around the @xmath17axis by @xmath15 . for both of geometries ( i ) and ( ii )
, we consider a rotation only in the right superconductor and the crystallographic @xmath14axis in the left half - space is selected parallel to the partition between the superconductors ( @xmath18axis ) . ]
the organization of the rest of this paper is as follows . in sec.[2 ]
the quasiclassical equations for green s functions are presented .
green s functions are obtained in sec.[3 ] .
the obtained formulas for the green s functions are used to calculate the charge and spin current densities and also density of states in the normal region . in sec . [ 4 ] , numerical results of charge and spin currents in the normal metal between superconductors are analyzed .
effects of normal layer thickness between superconductors and misorientation between @xmath11-vectors of superconductors on the charge and spin currents , and density of states have been investigated .
the paper will be finished with the conclusions in sec.[5 ] .
in this section , we consider a clean normal metal such as copper between two misorientated @xmath0wave superconductors .
a normal metal layer with the thickness of @xmath19 is sandwiched by two triplet superconductors .
interfaces of normal metal and superconductors have been considered as totally transparent . for the case
@xmath20 we can use the ballistic quasiclassical eilenberger equation @xcite @xmath21 = 0 , \label{eilenberger}\ ] ] and the normalization condition @xmath22 , where @xmath23 are discrete matsubara energies with @xmath24 , @xmath25 is the temperature , @xmath26 is the fermi velocity and @xmath27 in which @xmath28 is the pauli matrix in particle - hole space . @xmath29
denote pauli matrices in spin space in the following .
component ) of the charge current versus the phase difference @xmath30 for the axial state ( [ axial ] ) , @xmath31 , different thicknesses of normal metal and different misorientations between superconductors .
panel ( a ) is for geometry ( i ) and panel ( b ) for geometry ( ii ) .
currents are given in units of @xmath32 . ]
the matsubara propagator @xmath33 can be written in the standard form : @xmath34 [ green s function ] where the matrix structure of the off - diagonal self energy @xmath35 in the nambu space is @xmath36 [ order parameter ] in this paper , we focus on the unitary states , ( @xmath37 ) .
also , we use the eilenberger equation for @xmath38 in the normal metal region ( @xmath39 ) . from the eilenberger equation
it is clear that green s functions in the normal metal , @xmath40 and @xmath41 , are constant for @xmath42 .
solutions of eq.([eilenberger ] ) has to satisfy the conditions for green s functions at the bulk of superconductors @xmath43 .
component ) of the charge current versus the phase difference @xmath30 for the planar state ( [ planar ] ) , @xmath31 , different thicknesses of layer normal metal between superconductors and different misorientations between superconductors .
panel ( a ) is for geometry ( i ) and panel ( b ) for geometry ( ii ) . ] versus the phase difference @xmath30 for the axial state ( [ axial ] ) , geometry ( i ) and the different thickness of metal and different misorientation , @xmath31 and the different thicknesses of layer normal metal between superconductors .
spin currents @xmath44 for geometry ( i ) and @xmath45 for geometry ( ii ) are absent because of eq.([spin ] ) .
currents are calculated in units of @xmath46 . ] in addition , solutions of eqs .
( [ eilenberger ] ) satisfy the continuity conditions at the interfaces between metal and superconductors ( @xmath47 ) for all quasiparticle trajectories .
+ here , as in the paper @xcite , a simple step - like non - self - consistent model of the constant order parameter up to the interfaces is considered : @xmath48 where @xmath49 is the external phase difference between the gap functions of superconducting bulks .
we assume that the order parameter does not depend on coordinates and in each superconductors equals to its value far from the interface in the left or right bulks .
for such a model , the current - phase dependence of a josephson junction can be calculated for specific model of @xmath0wave pairing symmetry .
we believe that under this assumptions our results describe the real situation qualitatively@xcite . in the framework of such model
, the analytical expressions for the charge and spin currents and density of states can be obtained for an arbitrary form of the gap vector .
the solution of eilenberger equations allows us to calculate the charge and spin current densities in normal metal .
the expression for the charge current is : @xmath50 that for the spin current is : @xmath51 and that for local density of states at the energy e is : @xmath52 where @xmath53 stands for averaging over the directions of electron momentum on the fermi surface @xmath54 , @xmath55 is the electron density of states at the fermi surface and @xmath56 . versus the phase difference @xmath30 for the planar state ( [ planar ] ) , geometry ( i ) , @xmath31 , different thicknesses of layer normal metal between superconductors and different misorientations between superconductors .
note that spin currents @xmath57 and @xmath58 vanish . ] versus the phase difference @xmath30 for the planar state ( [ planar ] ) , geometry ( ii ) , @xmath31 , different thicknesses of layer normal metal between superconductors and different misorientations between superconductors .
spin currents of @xmath59 and @xmath60 are absent . ] as a function of misorientation angle for axial state ( [ axial ] ) , @xmath61 , @xmath62 .
solid lines are for geometry ( ii ) and dotted lines are for geometry ( i ) . here and in fig.[fig8 ] , we define @xmath63 and @xmath64 . ] ) , @xmath61 , @xmath62 .
solid lines are for geometry ( ii ) and dotted lines are for geometry ( i ) . ] , @xmath61 , @xmath62 and for both symmetries ( axial and planar states ) and different phases @xmath65 . ] , @xmath66 , @xmath62 , and different thicknesses of normal metal @xmath19 .
panels ( a ) and ( b ) are for axial state ( [ axial ] ) and panels ( c ) and ( d ) are for planar state ( [ planar ] ) . left plots are for geometry ( i ) and right plots are for geometry ( ii ) . ] , @xmath62 , @xmath66 , and two different misorientations .
panels ( a ) and ( b ) are for axial state ( [ axial ] ) and panels ( c ) and ( d ) for planar state ( [ planar ] ) . left plots are for geometry ( i ) and right plots are for geometry ( ii ) . in panel
( b ) , plots of @xmath67 and @xmath68 ( which are coincident ) are shown by a solid line . ]
the calculated green s functions in the normal metal are the following : @xmath69 @xmath70 where @xmath71 , @xmath72 , @xmath73 and @xmath74 using these green s function , we obtain the charge and spin current through the sns josephson junction and also the density of states of system in the normal region @xmath39 .
in this section we calculate charge and spin currents in sns junctions having in mind upt@xmath5 that is a @xmath0wave superconductor .
two famous models for order parameter of upt@xmath5 are considered .
the first is the axial state @xcite : @xmath75 and the other is the planar state @xcite : @xmath76 . \label{planar}\ ] ] the coordinate axes @xmath77 are chosen along the crystallographic axes @xmath78 in the left side of fig.[fig1 ] .
the function @xmath79 in eqs.([axial ] ) and ( [ planar ] ) describes the dependence of the order parameter @xmath80 on the temperature @xmath25 .
using green s function in eqs.([charge ] ) and ( [ spin ] ) , we have calculated charge and spin currents , and the density of states in @xmath0wave josephson junction with pairing symmetry of eqs .
( [ axial ] ) and ( [ planar ] ) numerically .
+ two superconducting bulks may have misorientation which has been created by geometries ( i ) and ( ii ) in fig .
[ fig1 ] . for two geometries and specific models of @xmath0wave pairing symmetry
we have plotted the currents in terms of phase difference in figs.[fig2]-[fig6 ] .
the obtained currents are periodic functions of phase difference between superconductors with a period of @xmath81 .
it is clarified that by increasing the thickness of normal layer , amplitude of currents decreases , which is understandable because by increasing thickness of the normal metal , quantum coherency of macroscopic phases between left and right superconductors decreases .
the current - phase relations are qualitatively different from the case of @xmath82wave @xcite and @xmath83wave josephson junction @xcite . in particular , an interesting case is the zero of charge current at finite phase @xmath84 in fig.[fig2 ] ( b ) which occurs for the most of junctions exactly at @xmath85 or @xmath86 . on the other hand , for finite misorientation between left and
right gap vectors even at the zero phase difference we obtain finite charge current . this is because according to eq .
( [ axial ] ) in geometry ( ii ) misorientation @xmath15 plays role of phase difference rotation by @xmath15 gives the phase factor @xmath87 .
also , in figs .
[ fig4]-[fig6 ] , it is found that the spin current of josephson junction between two triplet superconductors has been created by misorientation between gap vectors of superconductors @xcite .
for the case of geometry ( ii ) and axial state ( [ axial ] ) , spin current is absent because both of gap vectors are in the same direction .
also , by using eq .
( [ spin ] ) , we find that spin currents @xmath57 and @xmath58 for the axial state ( 12 ) and also the planar state ( 13 ) in geometry ( i ) , and spin currents @xmath59 and @xmath88 for the planar state ( 13 ) in geometry ( ii ) are absent .
in contrast to the charge current , the spin current has a large value at @xmath86 ( or at @xmath85 for large misorientation angle ) .
it is seen that charge current are generally odd function but spin current are even function of the phase difference because charge current is odd in time reversal while spin current is even . since spin
current is an even function of phase , its derivative becomes odd in phase and hence should be zero at @xmath85 and @xmath1 .
thus , the spin current has a local maximum or minimum at @xmath85 and @xmath1 .
we also see that for some phases as @xmath86 the spin current exists but charge current disappears .
this is a purely spin transport in the absence of charge current .
this effect can not be observed in singlet superconducting junction like conventional or high @xmath89 superconductors . only in the case of superconductor - ferromagnet - superconductor junction with inhomogeneous magnetization
this spin current is present @xcite .
the critical currents as a function of the misorientation angle are shown in figs.[fig7 ] and [ fig8 ] .
critical charge and spin currents , defined as @xmath63 and @xmath64 , respectively , are plotted as a function of misorientation angle @xmath15 for specified temperature and normal layer thickness ( @xmath90 , @xmath91 ) . by changing the misorientation
, we see the @xmath92 transition due to the anisotropy of the gap.@xcite also , at the @xmath92 transition point for both symmetries in geometry ( i ) we have a jump of spin current due to the jump of the phase as found in ref.@xcite but in geometry ( ii ) there are nt any @xmath92 transition and hence jump of spin current .
finally , in figs.[fig9]-[fig11 ] we have plotted the density of states in the normal region .
it is known that proximity effect in the sns junction strongly changes the density of states in the normal metal . in our calculations , we obtained andreev bound states in the junctions as shown in figs.[fig9]-[fig11 ] .
this is understandable , by interference effect of electrons and holes in the normal metal .
an electron will be reflected by andreev reflection as a hole .
this process is also repeated on the other sn interface .
then , interference of holes and electrons increases the effective andreev reflection probability and proximity effect at the low energies . in fig.[fig9 ] , we obtain a mini - gap like @xmath82wave and @xmath83wave superconductors @xcite .
exactly at @xmath86 we obtain a peak at the fermi energy . here , like josephson junction between other types of unconventional superconductors @xcite , we obtain the finite density of states for energies inside the gap ( @xmath93 ) .
the results of a finite misorientation and different @xmath19 are shown in fig .
[ fig10 ] .
the density of states gets smeared by increasing @xmath19 . in fig .
[ fig11 ] we plotted the density of states for both of geometries and both of symmetries .
it is observed that by increasing the misorientation two peaks are gradually merged into one and there will be a peak at zero energy at @xmath94 .
the increase of the density of states at low energy corresponds to that of the spin current .
we find that effect of thickness of normal layer is only quantitative but misorientation changes currents and the density of states qualitatively .
in this paper , we have investigated transport properties and also local density of states in triplet sns josephson junction with two @xmath0wave superconductors .
a spin - polarized current normal to the interface is investigated theoretically .
it is observed that normal layer decreases both of charge and spin currents .
it is also found that misorientation of @xmath11-vector between left and right superconducting bulks produces the spin - polarized current which can flow even in the absence of charge current .
we also unveiled a jump of the spin current associated with the 0-@xmath1 transition .
in addition , density of states have been calculated for various phase difference , normal metal thickness and misorientation .
normal metal thickness has only quantitative effect on the mid - gap states .
misorientation of the @xmath11-vectors drastically changes the junction property .
a. p. mackenzie and y. maeno , rev .
phys . * 75 * , 657 ( 2003 ) .
k. d. nelson , z. q. mao , y. maeno and y. liu , science * 306 * , 1151 ( 2004 ) .
y. asano , y. tanaka , m. sigrist and s. kashiwaya , phys .
b * 67 * , 184505 ( 2003 ) ; phys . rev .
b * 71 * , 214501 ( 2005 ) .
h. tou , y. kitaoka , k. ishida , k. asayama , n. kimura , y. onuki , e. yamamoto , y. haga and k. maezawa , phys .
lett . * 80 * , 3129 ( 1998 ) . v. muller , ch .
roth , d. maurer , e. w. scheidt , k. lers , e. bucher and h. e. bmel , phys .
lett . * 58 * , 1224 ( 1987 ) . y. j. qian , m. f. xu , a. schenstrom , h. p. baum , j. b. ketterson , d. hinks , m. levy and b. k. sarma , solid state commun .
, * 63 * , 599 ( 1987 ) .
a. a. abrikosov , j. of low temp . phys . * 53 * , 359 ( 1983 ) . h. fukuyama and y. hasegawa , j. phys .
jpn * 56 * , 877 ( 1987 ) .
a. g. lebed , k. machida and m. ozaki , phys .
b * 62 * , 795 ( 2000 ) .
s. s. saxena , p. agarwal , k. ahilan , f. m. grosche , r. k. w. haselwimmer , m. j. steiner , e. pugh , i. r. walker , s. r. julian , p. monthoux , g. g. lonzarich , a. huxley , i. shelkin , d. braithwaite , and j. flouquet , nature ( london ) * 406 * , 587 ( 2000 ) .
m. j. graf , s. k. yip and j. a. sauls , phys . rev .
b , * 62 * , 14393 ( 2000 ) .
k. machida , t. nishira , and t. ohmi , j. phys .
jpn , * 68 * , 3364 ( 1999 ) . b. lussier , b. ellman and l. taillefer , phys .
b , * 53 * , 5145 ( 1996 ) .
r. h. heffner and m. r. norman , comment on cond .
* 17 * , 361 ( 1996 ) .
y. tanaka , y. v. nazarov , and s. kashiwaya , phys .
lett . * 90 * , 167003 ( 2003 ) ; y. tanaka , yu .
v. nazarov , a. a. golubov , and s. kashiwaya , phys .
b * 69 * , 144519 ( 2004 ) .
y. tanaka , s. kashiwaya , and t. yokoyama , phys .
b * 71 * , 094513 ( 2005 ) .
y. tanuma , y. tanaka , and s. kashiwaya , phys .
b * 74 * , 024506 ( 2006 ) . g. d. guttman , e. b. jacob and d. j. bergman , phys . rev .
b * 57 * , 2717 ( 1998 ) .
r. lindell , j. penttila , m. sillanpaa and p. hakonen , phys . rev .
b * 68 * , 052506 ( 2003 ) .
a. a. golubov , m. yu . kupriyanov and e. ilichev , rev .
phys * 76 * , 411 ( 2004 ) .
g. rashedi and yu .
a. kolesnichenko , physica c , * 451 * , 31 ( 2007 ) . i. o. kulik and a. n. omelyanchouk , fiz .
temp * 4 * , 296 ( 1978 ) [ sov .
j. low temp .
phys . , * 4 * , 142 ( 1978 ) ] .
g. rashedi and yu .
kolesnichenko , phys .
b * 69 * , 024516 ( 2004 ) .
m. h. s. amin , m. coury , s. n. rashkeev , a. n. omelyanchouk and a. m. zagoskin , physica b * 318 * , 162 ( 2002 ) . k. k. likharev , rev .
phys * 51 * , 101 ( 1979 ) .
g. eilenberger , z. phys , * 214 * , 195 ( 1968 ) .
z. faraii and m. zareyan , phys .
b * 69 * , 014508 ( 2004 ) .
j. g. matthias , s. k. yip and j. a. sauls , phys . rev .
b * 62 * , 14393 ( 2000 ) .
m. alidoust , j. linder , g. rashedi , t. yokoyama and asle sudbo , phys .
b * 81 * , 014512 ( 2010 ) .
t. lofwander , v. s. shumeiko and g. wendin , supercond .
* 14 * , r53 ( 2001 ) . | we study transport properties and also local density of states in a clean superconductor - normal metal - superconductor josephson junction with triplet @xmath0wave superconductors based on the eilenberger equation .
effects of thickness of normal metal and misorientation between gap vector of superconductors on the spin and charge currents are investigated .
spin current , which stems from the misalignment of d - vectors , in the absence of charge current for some values of phase difference between superconductors is found .
we also find unconventional behavior of the spin current associated with 0-@xmath1 transition .
the misalignment of d - vectors also gives rise to a zero energy peak in the density of states in the normal metal . |
in stochastic analysis for diffusion processes , the bismut formula @xcite ( also known as bismut elworthy li formula due to @xcite ) and the integration by parts formula are two fundamental tools .
let , for instance , @xmath0 be the ( nonexplosive ) diffusion process generated by an elliptic differential operator on a riemannian manifold @xmath1 , and let @xmath2 be the associated markov semigroup . for @xmath3 and @xmath4 ,
the bismut formula is of type @xmath5 where @xmath6 is the diffusion process starting at point @xmath7 , @xmath8 is a random variable independent of @xmath9 and @xmath10 is the directional derivative along @xmath11 .
when the curvature of the diffusion operator is bounded below , this formula is available with @xmath8 explicitly given by @xmath11 and the curvature operator .
there exist a number of applications of this formula , in particular , letting @xmath12 be the density ( or heat kernel ) of @xmath2 w.r.t .
a nice reference measure @xmath13 , we have , formally , @xmath14 from ( [ 1.1 ] ) one may also derive gradient - entropy estimates of @xmath2 and thus , the following harnack inequality introduced in @xcite ( see @xcite ) : @xmath15\\[-10pt ] \eqntext{t>0 , p>1 , x , y\in m , f\in b_b(m),}\end{aligned}\ ] ] where @xmath16 is determined by moments of @xmath17 and thus , independent of @xmath9 .
this type of harnack inequality is a powerful tool in the study of contractivity properties , functional inequalities and heat kernel estimates ; see , for example , @xcite and references within . on the other hand , to characterize the derivative of @xmath12 in @xmath18 , which is essentially different from that in @xmath7 when @xmath2 is not symmetric w.r.t .
@xmath13 , we need to establish the following integration by parts formula ( see @xcite ) : @xmath19 for a smooth vector field @xmath11 and some random variable @xmath20 . combining this formula with ( [ 1.1 ] ) ,
we are able to estimate the commutator @xmath21 which is important in the study of flow properties ; see , for example , @xcite .
similar to ( [ 1.1 ] ) , inequality ( [ 1.3 ] ) can be used to derive a formula for @xmath22 and the shift harnack inequality of type @xmath23 \bigr ) ( x ) \mathrm{e}^{c_p(t , x , y ) } , \nonumber\\[-10pt]\\[-10pt ] \eqntext{t>0 , p>1 , x , y\in m , f\in b_b(m),}\end{aligned}\ ] ] where @xmath24 , is the exponential map on the riemannian manifold . differently from usual harnack inequalities like ( [ 1.2 ] ) , in ( [ 1.4 ] ) the reference function @xmath9 , rather than the initial point , is shifted .
this inequality will lead to different heat kernel estimates from known ones implied by ( [ 1.2 ] ) . before moving on ,
let us make a brief comment concerning the study of these two formulas .
the bismut formula ( [ 1.1 ] ) has been widely studied using both malliavin calculus and coupling argument ; cf . @xcite and references within .
although ( [ 1.3 ] ) also has strong potential of applications , it is , however , much less known in the literature due to the lack of efficient tools . to see that ( [ 1.3 ] )
is harder to derive than ( [ 1.1 ] ) , let us come back to @xcite where an explicit version of ( [ 1.3 ] ) is established for the brownian motion on a compact riemannian manifold . unlike the bismut formula which only relies on the ricci curvature , driver s integration by parts formula involves both the ricci curvature and its derivatives .
therefore , one can imagine that in general ( [ 1.3 ] ) is more complicated ( and hence harder to derive ) than ( [ 1.1 ] ) . to establish the integration by parts formula and the corresponding shift harnack inequality in a general framework , in this paper we propose a new coupling argument .
in contrast to usual coupling arguments where two marginal processes start from different points and meet at some time ( called the coupling time ) , for the new - type coupling the marginal processes start from the same point , but their difference reaches a fixed quantity at a given time . in the next section , we will introduce some general results and applications on the integration by parts formula and the shift harnack inequality using the new coupling method . the general result obtained in section [ sec2 ]
will be then applied in section [ sec3 ] to a class of degenerate diffusion processes , in section [ sec4 ] to delayed sdes and in section [ sec5 ] to semi - linear spdes .
we remark that the model considered in section [ sec3 ] goes back to the stochastic hamiltonian system , for which the bismut formula and the harnack inequalities have been investigated in @xcite by using both coupling and malliavin calculus .
as will be shown in section [ sec2.1 ] with a simple example of this model , for the study of the integration by parts formula and the shift harnack inequalities , the malliavin calculus can be less efficient than the new coupling argument .
in section [ sec2.1 ] we first recall the argument of coupling by change of measure introduced in @xcite for the harnack inequality and the bismut formula , and then explain how can we modify the coupling in order to derive the integration by parts formula and the shift harnack inequality , and introduce the malliavin calculus for the study of the integration by parts formula . in the second subsection we present some applications of the integration by parts formula and the shift harnack inequalities to estimates of the heat kernel and its derivatives . for a measurable space @xmath25 ,
let @xmath26 be the class of all bounded measurable functions on @xmath27 , and @xmath28 the set of all nonnegative elements in @xmath26 .
when @xmath27 is a topology space , we always take @xmath29 to be the borel @xmath30-field , and let @xmath31 [ resp .
, @xmath32 be the set of all bounded ( compactly supported ) continuous functions on @xmath27 .
if , moreover , @xmath27 is equipped with a differential structure , for any @xmath33 let @xmath34 be the set of all elements in @xmath31 with bounded continuous derivatives up to order @xmath35 , and let @xmath36 . finally , a contraction linear operator @xmath37 on @xmath26 is called a markov operator if it is positivity - preserving with @xmath38 .
let @xmath13 and @xmath39 be two probability measures on a measurable space @xmath25 , and let @xmath40 be two @xmath27-valued random variables w.r.t . a probability space @xmath41 . if the distribution of @xmath42 is @xmath13 , while under another probability measure @xmath43 on @xmath44 the distribution of @xmath45 is @xmath46 , we call @xmath47 a coupling by change of measure for @xmath13 and @xmath39 with changed probability @xmath48 . if @xmath13 and @xmath39 are distributions of two stochastic processes with path space @xmath27 , a coupling by change of measure for @xmath49 and @xmath39 is also called a coupling by change of measure for these processes . in this case @xmath42 and @xmath45 are called the marginal processes of the coupling . now , for fixed @xmath50 , consider the path space @xmath51}$ ] for some @xmath50 equipped with the product @xmath30-field @xmath52}$ ] .
let @xmath53 be a transition probability such that @xmath54 .
for any @xmath55 $ ] , let @xmath56 be the marginal distribution of @xmath57 at time @xmath58 .
then @xmath59 gives rise to a family of markov operators @xmath60}$ ] on @xmath61 with @xmath62 . in order to establish the harnack inequality , for any two different points ,
one constructs a coupling by change of measure @xmath47 for @xmath57 and @xmath63 with changed probability @xmath64 such that @xmath65 .
then @xmath66 this implies a harnack inequality of type ( [ 1.2 ] ) if @xmath67 . to establish the bismut formula ,
let , for example , @xmath27 be a banach space , and @xmath68 .
one constructs a family of couplings by change of measure @xmath69 for @xmath70 and @xmath57 with changed probability @xmath71 such that @xmath72 $ ] .
then , if @xmath73 exists in @xmath74 , for any @xmath75 , we obtain @xmath76 therefore , the bismut formula ( [ 1.1 ] ) is derived . on the other hand , for the integration by parts formula and shift harnack inequality
we need to construct couplings with marginal processes starting from the same point but their `` difference '' equals to a fixed value at time @xmath77 . for simplicity , below we only consider @xmath27 being a banach space . to extend the result to nonlinear spaces like riemannian manifolds
, one would need to make proper modifications using the geometric structure in place of the linear structure .
[ t2.1 ] let @xmath27 be a banach space and @xmath68 and @xmath50 be fixed . for any coupling by change of measure @xmath47 for @xmath57 and @xmath57 with changed probability @xmath78 such that @xmath79 , there holds the shift harnack inequality @xmath80 and the shift log - harnack inequality @xmath81 let @xmath82 $ ] , be a family of couplings by change of measure for @xmath57 and @xmath57 with changed probability @xmath83 such that @xmath84.\ ] ] if @xmath85 and @xmath86 the proof is similar to that introduced above for the harnack inequality and the bismut formula .
note that @xmath87 .
we have @xmath88 next , by the young inequality ( see @xcite , lemma 2.4 ) , for positive @xmath9 we have @xmath89 noting that @xmath90 , we obtain @xmath91 provided @xmath85 and @xmath92 exists in @xmath74 . from theorem [ t2.1 ] and its proof we see that the machinery of the new coupling argument is very clear .
so , in applications the key point of the study lies in the construction of new type couplings .
next , we explain how one can establish the integration by parts formula using malliavin calculus .
let , for example , @xmath93 be the cylindrical brownian motion on an hilbert space @xmath94 w.r.t . a probability space @xmath41 with natural filtration @xmath95 .
let @xmath96;h \bigr)\dvtx \|h \|_{h^1}^2:=\int_0^t \bigl|h'(s ) \bigr|^2\,\d s<\infty \biggr\}\ ] ] be the cameron martin space .
for a measurable functional of @xmath97 , denoted by @xmath98 , such that @xmath99 and @xmath100 gives rise to a bounded linear operator .
then we write @xmath101 and call @xmath102 the malliavin gradient of @xmath98 .
it is well known that @xmath103 is a densely defined closed operator on @xmath104 ; see , for example , @xcite , section 1.3 .
let @xmath105 be its adjoint operator , which is also called the divergence operator .
[ ma ] let @xmath106 and @xmath107 be introduced above .
let @xmath108 and @xmath109 . if there exists @xmath110 such that @xmath111 , then @xmath112 since @xmath111 , we have @xmath113 finally , as the integration by parts formula ( [ iit ] ) and by the young inequality ( see @xcite , lemma 2.4 ) imply the derivative - entropy inequality @xmath114 \biggr\}p_tf,\qquad \delta>0\end{aligned}\ ] ] and the @xmath115-derivative inequality
@xmath116 according to the following result it also implies shift harnack inequalities .
[ pis ] let @xmath37 be a markov operator on @xmath26 for some banach space @xmath27 .
let @xmath117 .
let @xmath118 and @xmath119 .
then @xmath120 holds for any positive @xmath121 if and only if @xmath122\\[-14pt ] & & { } \times\exp \biggl[\int_0 ^ 1 \frac{pr}{1+(p-1)s } \beta_e \biggl(\frac{p-1}{r+r ( p-1)s } , \cdot+sre \biggr ) \,\d s \biggr]\nonumber\end{aligned}\ ] ] holds for any positive @xmath123 and @xmath124 .
let @xmath125 be a constant .
then @xmath126 is equivalent to @xmath127 the proof of ( 1 ) is similar to that of @xcite , proposition 4.1 , while ( 2 ) is comparable to @xcite , proposition 1.3 . let @xmath128 $ ] . by the monotone class theorem , it suffices to prove for @xmath129 . since @xmath130 for @xmath131 , it follows from ( [ w12ba ] ) that @xmath132 \\ & & \hspace*{173pt}{}+ prp\bigl\{\nabla_ef^{\beta(s)}(sre+ \cdot)\bigr\ } \bigr ) ( x ) \\ & & \qquad{}\ge- \frac{rp}{\beta(s)}\beta_e \biggl(\frac{p-1}{r\beta ( s ) } , x+sre \biggr),\qquad s\in[0,1].\end{aligned}\ ] ] taking the integral over @xmath133 $ ] w.r.t .
@xmath134 we prove ( [ w12bb ] ) .
next , let @xmath135 be fixed , and assume that @xmath136 ( otherwise , simply use @xmath137 to replace @xmath138 ) .
then ( [ w12bb ] ) with @xmath139 implies that @xmath140-pf(z)\biggr\ } \\ & & \qquad= \delta p(f\log f)(z ) + \beta_e(\delta ) pf(z).\end{aligned}\]]therefore , ( [ w12ba ] ) holds .
let @xmath141 . for nonnegative @xmath129 , ( [ l2 g
] ) implies that @xmath142 noting that @xmath143 we obtain @xmath144 minimizing the right - hand side in @xmath141 , we prove ( [ sht ] ) . on the other hand ,
let @xmath145 .
without loss of generality we assume that @xmath146 , otherwise it suffices to replace @xmath138 by @xmath137
. then ( [ sht ] ) implies that @xmath147 therefore , ( [ l2 g ] ) holds . to conclude this section
, we would like to compare the new coupling argument with known coupling arguments and the malliavin calculus , from which we see that the study of the integration by parts formula and the shift harnack inequality is , in general , more difficult than that of the bismut formula and the harnack inequality . first , when a strong markov process is concerned , for a usual coupling @xmath148 one may ask that the two marginal processes move together after the coupling time , so that to ensure @xmath65 , one only has to confirm that the coupling time is not larger than the given time @xmath77 . but for the new coupling argument , we have to prove that at time @xmath77 , the difference of the marginal processes equals to a fixed quantity , which can not be ensured , even if the difference already reached this quantity at a ( random ) time before @xmath77 . from this
we see that construction of a new - type coupling is , in general , more difficult than that of a usual coupling .
second , it is well known that the malliavin calculus is a very efficient tool to establish bismut - type formulas . to see the difficulty for deriving the integration by parts formula using malliavin calculus , we look at a simple example of the model considered in section [ sec3 ] , that is , @xmath149 is the solution to the following degenerate stochastic equation on @xmath150 : @xmath151 where @xmath152 is the one - dimensional brownian motion and @xmath153 . for this model
the bismut formula and harnack inequalities can be easily derived from both the coupling method and malliavin calculus ; see @xcite .
we now explain how can one establish the integration by parts formula using malliavin calculus . for fixed @xmath50 and , for example , @xmath154 , to derive the integration by parts formula for the derivative along @xmath138 using theorem [ ma ]
, one needs to find @xmath110 such that @xmath155 to search for such an element @xmath156 , we note that ( [ ee ] ) implies @xmath157 and @xmath158 where @xmath159 then , ( [ dh ] ) is equivalent to @xmath160 it is , however , very hard to solve @xmath156 from this equation for general @xmath161 . on the other hand , we will see in section [ sec3 ] that the coupling argument we proposed above is much more convenient for deriving the integration parts formula for this example .
we first consider @xmath162 for some @xmath163 , and to estimate the density w.r.t .
the lebesgue measure for distributions and markov operators using integration by parts formulas and shift harnack inequalities .
[ t4.1 ] let @xmath42 be a random variable on @xmath164 such that for some @xmath165 @xmath166 the distribution @xmath167 of @xmath42 has a density @xmath168 w.r.t .
the lebesgue measure , which satisfies @xmath169 consequently , for any @xmath170 and any convex positive function @xmath171 , @xmath172 for any @xmath173 , @xmath174 \(1 ) we first observe that if @xmath167 has density @xmath168 , then for any @xmath175 , @xmath176 this implies ( [ dd0 ] ) . to prove the existence of @xmath168 ,
let @xmath177 be the distribution density function of @xmath178 , where @xmath179 is the standard gaussian random variable on @xmath164 independent of @xmath42 and @xmath180 .
it follows from ( [ 4.1 ] ) that @xmath181 then @xmath182 so , the sequence @xmath183 is bounded in @xmath184 .
thus , up to a subsequence , @xmath185 in @xmath186 for some nonnegative function @xmath168 . on the other hand , we have @xmath187 weakly . therefore , @xmath188 .
\(2 ) as for the second assertion , noting that for @xmath189 one has @xmath190 it follows from ( [ 4.1 ] ) that @xmath191 next , we consider applications of a general version of the shift harnack .
let @xmath192 be a transition probability on a banach space @xmath27 .
let @xmath193 be the associated markov operator .
let @xmath194 be a strictly increasing and convex continuous function .
consider the shift harnack inequality @xmath195 for some @xmath196 and constant @xmath197 . obviously ,
if @xmath198 for some @xmath199 , then this inequality reduces to the shift harnack inequality with power @xmath200 , while when @xmath201 , it becomes the log shift harnack inequality .
[ t4.2 ] let @xmath37 be given above and satisfy ( [ ph ] ) for all @xmath202 and some nonnegative measurable function @xmath203 on @xmath204 .
then @xmath205 consequently : if @xmath206 , then @xmath37 has a transition density @xmath207 w.r.t .
the lebesgue measure such that @xmath208 if @xmath198 for some @xmath199 , then @xmath209 let @xmath210 such that @xmath211 . by ( [ ph ] )
we have @xmath212 integrating both sides w.r.t . @xmath213 and noting that @xmath214 , we obtain @xmath215 this implies ( [ aa0 ] ) .
when @xmath206 , ( [ aa0 ] ) implies that @xmath216 since by the strictly increasing and convex properties we have @xmath217 as @xmath218 .
now , for any lebesgue - null set @xmath219 , taking @xmath220 we obtain from @xmath206 that @xmath221 therefore , applying ( [ aa2 ] ) to @xmath222 we obtain @xmath223 which goes to zero as @xmath224 .
thus @xmath225 is absolutely continuous w.r.t .
the lebesgue measure , so that the density function @xmath207 exists , and ( [ aa1 ] ) follows from ( [ aa0 ] ) by taking @xmath226 .
finally , let @xmath198 for some @xmath199 .
for fixed @xmath7 , let @xmath227 it is easy to see that @xmath228 .
then it follows from ( [ aa0 ] ) with @xmath198 that @xmath229 then ( [ aa ] ) follows by letting @xmath224 . finally , we consider applications of the shift harnack inequality to distribution properties of the underlying transition probability .
[ t4.3 ] let @xmath37 be given above for some banach space @xmath27 , and let ( [ ph ] ) hold for some @xmath68 , finite constant @xmath230 and some strictly increasing and convex continuous function @xmath231 .
@xmath225 is absolutely continuous w.r.t . @xmath232 . if @xmath233 for some strictly increasing positive continuous function @xmath234 on @xmath235 .
then the density @xmath236 satisfies @xmath237 for @xmath232-null set @xmath219 , let @xmath238 .
then ( [ ph ] ) implies that @xmath239 , hence @xmath240 since @xmath241 for @xmath141 .
therefore , @xmath225 is absolutely continuous w.r.t .
@xmath232 . next ,
let @xmath233 . applying ( [ ph ] ) for @xmath242 and noting that @xmath243 we obtain @xmath244
then the proof is complete by letting @xmath224 .
consider the following degenerate stochastic differential equation on @xmath245 @xmath246 : @xmath247 where @xmath219 and @xmath248 are two matrices of order @xmath249 and @xmath250 , respectively , @xmath251 is measurable with @xmath252 for @xmath253 , @xmath254 are invertible @xmath255-matrices measurable in @xmath58 such that the operator norm @xmath256 is locally bounded and @xmath152 is the @xmath257-dimensional brownian motion .
when @xmath258 this equation is degenerate , and when @xmath259 we set @xmath260 , so that the first equation disappears and thus , the equation reduces to a nondegenerate equation on @xmath164 . to ensure the existence of the transition density ( or heat kernel ) of the associated semigroup @xmath2 w.r.t . the lebesgue measure on @xmath261 , we make use of the following kalman rank condition ( see @xcite ) which implies that the associated diffusion is subelliptic , @xmath262=m$.}\ ] ] when @xmath259 this condition is trivial , and for it means that that is , @xmath263 . for any @xmath264 and @xmath163
, there exist plenty of examples for matrices @xmath219 and @xmath248 such that holds ; see @xcite .
therefore , we allow that @xmath265 is much larger than @xmath257 , so that the associated diffusion process is highly degenerate ; see example [ exa3.1 ] below .
it is easy to see that if @xmath266 , @xmath248 is symmetric and @xmath267 for some smooth functions @xmath268 and @xmath269 , then ( [ 3.1 ] ) reduces to the hamiltonian system @xmath270 with hamiltonian function @xmath271 see , for example , @xcite . if , in particular , @xmath272 and @xmath273 for some constant @xmath274 , the corresponding fokker
planck equation is known as the `` kinetic fokker
planck equation '' in pde ( see @xcite ) , and the stochastic equation is called `` stochastic damping hamiltonian system '' ; see @xcite .
let the solution to ( [ 3.1 ] ) be nonexplosive , and let @xmath275 to state our main results , let us fix @xmath50 . for nonnegative @xmath276)$ ] with @xmath277 in @xmath278 , define @xmath279 then @xmath280 is invertible ; cf . @xcite . for
any @xmath281 and @xmath141 , let @xmath282 be the ball centered at @xmath283 with radius @xmath284 .
[ t3.1 ] assume and that the solution to ( [ 3.1 ] ) is nonexplosive such that @xmath285 } \mathbf e \bigl\{\sup _ { b(x(t ) , y(t ) ; r ) } \bigl|\nabla z(t,\cdot ) \bigr|^2 \bigr\}<\infty,\qquad r>0.\ ] ] let @xmath286)$ ] such that @xmath287 in @xmath278 , and @xmath288 moreover , for @xmath289 , let @xmath290.\end{aligned}\ ] ] for any @xmath291 , there holds @xmath292 let @xmath293 and @xmath294,\ ] ] where @xmath295 with @xmath296 then @xmath297 we only prove ( 1 ) , since ( 2 ) follows from theorem [ t2.1 ] with the coupling constructed below for @xmath298 .
let @xmath299 solve ( [ 3.1 ] ) with initial data @xmath300 , and for @xmath301 $ ] let @xmath302 solve the equation @xmath303 then it is easy to see that @xmath304 combining this with @xmath305 and ( [ ps ] ) , we see that @xmath306 and @xmath307 therefore , @xmath308.\ ] ] next , to see that @xmath309 is a coupling by change of measure for the solution to ( [ 3.1 ] ) , reformulate ( [ ccc0 ] ) as @xmath310 where @xmath311.}\end{aligned}\ ] ] let @xmath312 and @xmath313.\ ] ] by lemma [ ll ] below and the girsanov theorem , @xmath314 is a @xmath257-dimensional brownian motion under the probability measure @xmath315 .
therefore , @xmath316 is a coupling by change of measure with changed probability @xmath317 .
moreover , combining ( [ sol ] ) with the definition of @xmath318 , we see from ( [ uu ] ) that @xmath319 holds in @xmath74
. then the proof is complete by theorem [ t2.1](2 ) .
[ ll ] let the solution to ( [ 3.1 ] ) be nonexplosive such that ( [ uu ] ) holds , and let @xmath320 be in ( [ xi ] ) . then for any @xmath321 $ ] the process @xmath322,\nonumber \\ \eqntext{t\in[0,t]}\end{aligned}\ ] ] is a uniformly integrable martingale with @xmath323}\mathbf e \ { r_\varepsilon(t)\log r_\varepsilon(t )
\}<\infty$ ] .
let @xmath324 .
then @xmath325 as .
it suffices to show that @xmath326,n\ge1}\mathbf e \bigl\ { r_\varepsilon(t \land \tau _ n)\log r_\varepsilon(t\land\tau_n ) \bigr \ } < \infty.\ ] ] by ( [ sol ] ) , there exists @xmath141 such that @xmath327,\varepsilon\in[0,1].\ ] ] let @xmath328 . by the girsanov theorem , @xmath329}$ ]
is the dimensional brownian motion under the changed probability @xmath330 .
then , due to ( [ qqq ] ) , @xmath331}\mathbf e \bigl\{r_\varepsilon(t\land \tau _
n)\log r_\varepsilon(t\land\tau_n ) \bigr\ } \\ & & \qquad = \frac1 2 \mathbf e_{\mathbf q_{\varepsilon , n } } \int_0^{t\land \tau_n } \bigl|\sigma(s)^{-1}\xi_\varepsilon(s ) \bigr|^2\,\d s \\ & & \qquad \le c + c \mathbf e_{\mathbf q_{\varepsilon , n}}\int_0^{t\land\tau _
n } \sup_{b(x^\varepsilon(t ) , y^\varepsilon ( t ) ; r ) } \bigl|\nabla z(t,\cdot ) \bigr|^2\,\d t\end{aligned}\ ] ] holds for some constant @xmath332 independent of @xmath333 .
since the law of @xmath334 under @xmath330 coincides with that of @xmath335 under @xmath336 , combining this with ( [ uu ] ) , we obtain @xmath331}\mathbf e
\bigl\{r_\varepsilon(t\land \tau_n)\log r_\varepsilon(t\land\tau_n ) \bigr\ } \\ & & \qquad \le c + c \int_0^{t } \mathbf e\sup _ { b(x(t ) , y(t ) ; r ) } \bigl|\nabla z(t,\cdot ) \bigr|^2\,\d t<\infty.\end{aligned}\ ] ] therefore , ( [ wwl ] ) holds .
[ rem3.1 ] ( a ) as shown in @xcite , lemma 2.4 , condition ( [ uu ] ) is implied by the lyapunov condition ( a ) therein , for which some concrete examples have been presented in @xcite . moreover , as shown in @xcite , section 3 ( see also theorem 4.1 in @xcite ) that under reasonable grown conditions of @xmath337 one obtains from theorem [ t3.1](1 ) @xmath338 for some constant @xmath339 and some positive functions @xmath340 . according to theorem [ ma ] , this inequality implies the shift harnack inequality .
\(b ) for any @xmath341 . applying theorem [ t3.1 ] to @xmath342 in place of @xmath149
, we see that the assertions in theorem [ t3.1 ] hold for @xmath343 in place of @xmath344 with @xmath77 and @xmath345 replaced by @xmath346 and @xmath347 , respectively . to derive explicit inequalities from theorem [ t3.1 ]
, we consider below a special case where @xmath348 is bounded and @xmath349 for some natural number @xmath350 .
[ c3.2 ] assume .
if @xmath351 and @xmath352 are bounded in @xmath253 , and @xmath349 for some @xmath350 . then there exists a constant @xmath332 such that for any positive @xmath353 and @xmath289 : @xmath354 $ ] , @xmath199 ; @xmath355 ; for @xmath291 , @xmath356 ; for strictly positive @xmath291 , @xmath357 according to remark [ rem3.1](b ) , @xmath358 and the jensen inequality , we only need to prove for @xmath359 $ ] .
. then @xmath305 and due to @xcite , theorem 4.2(1 ) , the rank condition implies that @xmath361 for some constant @xmath362 independent of @xmath359 $ ] . to fix the other reference function @xmath363 in theorem [ t3.1 ] ,
let @xmath364 be such that @xmath365 take @xmath366.\ ] ] then @xmath367 and @xmath368 for @xmath369 .
since @xmath349 , we conclude that @xmath370 .
therefore , ( [ ps ] ) holds .
it is easy to see that @xmath371\ ] ] holds for some constant @xmath362 .
combining this with ( [ q ] ) , ( [ sol ] ) and the boundedness of @xmath372 and @xmath373 , we obtain @xmath374 \\[-8pt ] \bigl|\theta(t ) \bigr|&\le & c
\bigl(t^{-(2k+1)}|e_1|+ |e_2| \bigr ) \nonumber\end{aligned}\ ] ] for some constant @xmath362 . from this and theorem [ t3.1 ]
, we derive the desired assertions . in the situation of corollary [ c3.2 ] .
let @xmath375 be the operator norm from @xmath376 to @xmath377 w.r.t .
the lebesgue measure on @xmath378
. then there exists a constant @xmath332 such that @xmath379\\[-12pt ] \eqntext{p>1 , t>0.}\end{aligned}\ ] ] consequently , the transition density @xmath380 of @xmath381 w.r.t .
the lebesgue measure on @xmath378 satisfies @xmath382 by corollary [ c3.2](1 ) , ( [ cc ] ) follows from ( [ aa0 ] ) for @xmath383 , @xmath198 and @xmath384 moreover , ( [ hhh ] ) follows from ( [ aa ] ) .
[ exa3.1 ] a simple example for theorem [ c3.2 ] to hold is that @xmath385 and @xmath386 are independent of @xmath58 with @xmath387 , @xmath388 and@xmath389 .
in this case we have @xmath390 ; that is , the dimension of the generate part is controlled by that of the nondegenerate part . in general , our results allow @xmath265 to be much larger than @xmath257 .
for instance , let @xmath391 for some @xmath392 and @xmath393 then @xmath349 and holds for @xmath394 .
therefore , assertions in corollary [ c3.2 ] hold for @xmath395 .
the purpose of this section is to establish driver s integration by parts formula and shift harnack inequality for delayed stochastic differential equations . in this case
the associated segment processes are functional - valued , and thus , infinite - dimensional . as continuation to section [ sec3 ] ,
it is natural for us to study the generalized stochastic hamiltonian system with delay as in @xcite , where the bismut formula and the harnack inequalities are derived using coupling .
however , for this model it seems very hard to construct the required new - type couplings .
so , we only consider here the nondegenerate setting .
let @xmath396 be a fixed number , and let @xmath397;\mathcal r^d)$ ] be equipped with uniform norm @xmath398 . for simplicity , we will use @xmath399 to denote the gradient operator both on @xmath164 and @xmath400 .
for instance , for a differentiable function @xmath269 on @xmath400 and @xmath401 , @xmath402 is a linear operator from @xmath400 to @xmath403 with @xmath404 moreover , let @xmath405 be the operator norm for linear operators . finally , for a function @xmath406 and @xmath253 , let @xmath407 be such that @xmath408 , @xmath409 $ ] .
consider the following stochastic differential equations on @xmath410 : @xmath411 where @xmath152 is the brownian motion on @xmath164 , @xmath412 is measurable such that @xmath413 is locally bounded in @xmath58 , and @xmath414 is measurable with @xmath352 locally bounded .
we remark that the local boundedness assumption of @xmath413 is made only for simplicity and can be weakened by some growth conditions as in @xcite .
now , for any @xmath415 , let @xmath416 be the solution to ( [ 4.1 ] ) for @xmath417 , and let @xmath418 be the associated segment process .
let @xmath419 we aim to establish the integration by parts formula and shift harnack inequality for @xmath381 .
it turns out that we are only able to make derivatives or shifts along directions in the cameron
martin space @xmath420 [ t3.1 ] let @xmath421 and @xmath422 be fixed .
for any @xmath423)$ ] such that @xmath424 , let @xmath425 $ , \vspace*{2pt } \cr \eta'(t - t ) , & \quad if $ t\in(t-\tau , t]$.}\ ] ] let @xmath426 for @xmath427 $ ] . for any @xmath428 , @xmath429 holds for @xmath430.\ ] ] consequently , for any @xmath431 and positive @xmath428 , @xmath432 for any nonnegative @xmath433 , @xmath434.\end{aligned}\ ] ] for any positive @xmath433 , @xmath435 for fixed @xmath415 , let @xmath0 solve ( [ 4.1 ] ) for @xmath417 .
for any @xmath321 $ ] , let @xmath436 solve the equation @xmath437 then it is easy to see that @xmath438.\ ] ] in particular , @xmath439 . next , let @xmath440.\end{aligned}\ ] ] by the girsanov theorem , under the changed probability @xmath441 , the process @xmath442\ ] ] is a @xmath257-dimensional brownian motion .
so , @xmath443 is a coupling by change of measure with changed probability @xmath444 .
then the desired integration by parts formula follows from theorem [ t2.1 ] since @xmath85 and due to ( [ th0 ] ) , @xmath445 holds in @xmath74 .
taking @xmath446 , we have @xmath447 then @xmath448
so , @xmath449 \\ & & \qquad\le\frac1 2 \log\mathbf e\exp \biggl [ \frac{2 k(t)^2}{\delta ^2 } \int _ 0^t \bigl|\gamma ( t)-\nabla_{\theta_t}b(t , \cdot ) ( x_t ) \bigr|^2\,\d t \biggr ] \\ & & \qquad\le\frac{2k(t)^2(1+t^2\kappa(t)^2)}{\delta^2 } \biggl(\|\eta\| _ \mathcal h^2 + \frac{|\eta(-\tau)|^2}{t-\tau } \biggr).\end{aligned}\ ] ] then the second result in ( 1 ) follows from the young inequality @xmath450.\end{aligned}\ ] ] finally , ( 2 ) and ( 3 ) can be easily derived by applying theorem [ t2.1 ] for the above constructed coupling with @xmath298 , and using ( [ th0 ] ) and ( [ llo ] )
. from theorem [ t3.1 ] we may easily derive regularization estimates on @xmath451 , the distribution of @xmath452 . for instance , theorem [ t3.1](1 ) implies estimates on the derivative of @xmath453 along @xmath422 for @xmath415 and measurable @xmath454 ; and due to theorems [ t4.3 ] , [ t3.1](2 ) and [ t3.1](3 ) imply some integral estimates on the density @xmath455 for @xmath456 .
moreover , since @xmath457 is dense in @xmath400 , the shift harnack inequality in theorem [ t3.1](2 ) implies that @xmath451 has full support on @xmath400 for any @xmath421 and @xmath415 .
the purpose of this section is to establish driver s integration by parts formula and shift harnack inequality for semi - linear stochastic partial differential equations .
we note that the bismut formula has been established in @xcite for a class of delayed spdes , but for technical reasons we only consider here the case without delay .
let @xmath94 be a real separable hilbert space , and @xmath458 a cylindrical wiener process on @xmath171 with respect to a complete probability space @xmath459 with the natural filtration @xmath460 .
let @xmath461 and @xmath462 be the spaces of all linear bounded operators and hilbert schmidt operators on @xmath171 , respectively .
denote by @xmath405 and @xmath463 the operator norm and the hilbert
schmidt norm , respectively .
consider the following semi - linear spde : @xmath464 where @xmath465 is a linear operator on @xmath171 generating a contractive , strongly continuous semigroup @xmath466 such that @xmath467 .
@xmath468 is measurable , and frchet differentiable in the second variable such that @xmath469 is locally bounded in @xmath253 .
@xmath470 is measurable and locally bounded , and @xmath471 is invertible such that @xmath472 is locally bounded in @xmath253 .
then the equation ( [ eq1 ] ) has a unique a mild solution ( see @xcite ) , which is an adapt process @xmath473 on @xmath171 such that @xmath474 let @xmath475 finally , for any @xmath108 , let @xmath476 [ t5.1 ] let @xmath50 and @xmath477 be fixed .
let @xmath478 , @xmath479 for @xmath427 $ ]
. for any @xmath480 , @xmath481 consequently , for any @xmath431 and positive @xmath480 , @xmath482 for any nonnegative @xmath483 , @xmath484.\ ] ] for any positive @xmath483 , @xmath485 for fixed @xmath486 , let @xmath0 solve ( [ 4.1 ] ) for @xmath487 . for any @xmath321 $ ] , let @xmath436 solve the equation @xmath488 then it is easy to see that @xmath489.\ ] ] in particular , @xmath490
. next , let @xmath491.\end{aligned}\ ] ] by the girsanov theorem , under the weighted probability @xmath441 , the process @xmath492\ ] ] is a @xmath257-dimensional brownian motion .
so , @xmath493 is a coupling by change of measure with changed probability @xmath317 .
then the desired integration by parts formula follows from theorem [ t2.1 ] since @xmath85 and due to ( [ th ] ) , @xmath494 holds in @xmath74 .
this formula implies the second inequality in ( 1 ) due to the given upper bounds on @xmath352 and @xmath495 and the fact that @xmath496p_tf \\ & & \qquad \le\frac\delta2 \log\mathbf e\exp \biggl[\frac2 { \delta^2 } \int_0^t \bigl|\sigma ( t)^{-1 } \bigl(e- \nabla_{e(t)}b(t,\cdot ) \bigl(x(t ) \bigr ) \bigr ) \bigr|^2\,\d t \biggr]p_tf.\end{aligned}\ ] ] finally , since @xmath497 , ( 2 ) and ( 3 ) can be easily derived by applying theorem [ t2.1 ] for the above constructed coupling with @xmath298 .
the author would like to thank the referees for helpful comments and corrections . | a new coupling argument is introduced to establish driver s integration by parts formula and shift harnack inequality . unlike known coupling methods where
two marginal processes with different starting points are constructed to move together as soon as possible , for the new - type coupling the two marginal processes start from the same point but their difference is aimed to reach a fixed quantity at a given time . besides the integration by parts formula , the new coupling method is also efficient to imply the shift harnack inequality . differently from known harnack inequalities where the values of a reference function at different points are compared , in the shift harnack inequality the reference function , rather than the initial point ,
is shifted .
a number of applications of the integration by parts and shift harnack inequality are presented .
the general results are illustrated by some concrete models including the stochastic hamiltonian system where the associated diffusion process can be highly degenerate , delayed sdes and semi - linear spdes . |
in @xcite , church , ellenberg , and farb studied representations of the category @xmath0 , consisting of finite sets with injective maps , and found numerous applications to topology and algebra . in @xcite , we studied representations of the closely related category @xmath1 ( in addition to many other examples ) , where @xmath2 is the category of finite sets with surjective maps , and also found several applications ( e.g. , to the lannes schwartz artinian conjecture and to @xmath3-modules ) . in this paper , we study generalizations of @xmath0 and @xmath1 in which a group @xmath4 has been added to the mix .
our main tool is the theory developed in @xcite ; in fact , the example @xmath5 was a primary source of motivation for that paper . in the rest of the introduction
, we summarize our motivations and results .
let @xmath4 be a group .
a * @xmath4-map * between finite sets @xmath6 and @xmath7 is a pair @xmath8 consisting of functions @xmath9 and @xmath10 .
if @xmath11 and @xmath12 are two @xmath4-maps , their composition @xmath13 is defined by @xmath14 and @xmath15 ( the product taken in @xmath4 ) .
we let @xmath16 be the category whose objects are finite sets and whose morphisms are @xmath4-maps .
we let @xmath17 ( resp .
@xmath18 ) be the subcategory where the function @xmath19 is injective ( resp .
surjective ) .
we note that the automorphism group of the set @xmath20=\{1 , \ldots , n\}$ ] in any of these categories is the wreath product @xmath21 .
thus a representation of any of these categories can be thought of as a sequence @xmath22 , where @xmath23 is a representation of @xmath21 , equipped with certain transition maps between @xmath23 and @xmath24 .
( the kind of transition maps depends on the category . )
these representations are the subject of this paper .
a @xmath4-map @xmath25 is the same as a @xmath4-equivariant map @xmath26 .
thus @xmath16 is equivalent to the category whose objects are free @xmath4-sets with finitely many orbits , and whose morphisms are @xmath4-equivariant functions . usually , noetherianity is the first property one wants to establish about the representation theory of a category .
( see [ ss : rep ] for the definition of `` noetherian '' in this context . ) for @xmath0 , this was proved in @xcite , in varying levels of generality .
our main result about @xmath17 characterizes when representations are noetherian : [ mainthm : fi1 ] let @xmath27 be a ring
. then @xmath28 is noetherian if and only if the group algebra @xmath29 $ ] is left - noetherian for all @xmath30 .
recall that a group @xmath4 is * polycyclic * if it has a finite composition series @xmath31 such that @xmath32 is cyclic for @xmath33 , and it is * polycyclic - by - finite * ( or * virtually polycyclic * ) if it contains a polycyclic subgroup of finite index
. it is known ( * ? ? ?
* , lemma 3 ) that the group ring of a polycyclic - by - finite group over a left - noetherian ring is left - noetherian ( there it is stated for the integral group ring , but the proof works for any left - noetherian coefficient ring ) .
in fact , there are no other known examples of noetherian group algebras , but see @xcite for related results . since a finite product of polycyclic - by - finite groups is again polycyclic - by - finite , the above theorem gives : [ cor : polycyclic ] let @xmath4 be a polycyclic - by - finite group and let @xmath27 be a left - noetherian ring .
then @xmath28 is noetherian . when @xmath4 is a finite group , we prove a stronger result : [ mainthm : fi2 ] if @xmath4 is finite then the category @xmath17 is quasi - grbner .
`` quasi - grbner '' is a purely combinatorial condition on a category , introduced in @xcite ( and recalled in [ ss : rep ] below ) , that implies noetherianity of the representation category .
thus theorem [ mainthm : fi2 ] implies theorem [ mainthm : fi1 ] when @xmath4 is finite , as stated .
however , quasi - grbner gives more than just noetherianity : it implies that representations admit a theory of grbner bases , in an appropriate sense , and thus computations with representations can be carried out algorithmically ( at least in principle ) .
the proof of theorem [ mainthm : fi2 ] is an easy consequence of the theory developed in @xcite .
the proof of theorem [ mainthm : fi1 ] is more involved .
the key input is the fact that @xmath34 is noetherian whenever @xmath35 is left - noetherian .
this is applied with @xmath36 $ ] , so even if one only cares about theorem [ mainthm : fi1 ] when @xmath27 is a field , the proof uses @xmath0-modules over non - commutative rings .
in fact , this is the first real application of @xmath0-modules over non - commutative rings that we know of .
there is one additional result on @xmath17-modules worth mentioning : [ mainthm : fi3 ] suppose that @xmath4 is finite and @xmath27 is a splitting field for @xmath4 in which the order of @xmath4 is invertible
. then representations of @xmath17 are equivalent to representations of @xmath37 , where @xmath38 is the number of non - trivial irreducible representations of @xmath4 over @xmath27 . here
@xmath39 is the groupoid of finite sets , i.e. , the category of finite sets with bijections as morphisms .
thus , in the context of the theorem , the theory of @xmath17-modules reduces to the theory of @xmath0-modules .
we only study @xmath18 when @xmath4 is a finite group .
the first result is about noetherianity , and is an easy consequence of the theory of @xcite : [ mainthm : fs1 ] if @xmath4 is finite then @xmath5 is quasi - grbner . in particular , if @xmath27 is left - noetherian , then @xmath40 is noetherian .
we next turn to hilbert series .
suppose that @xmath41 is an @xmath5-module over a field @xmath27 , and suppose that @xmath4 has @xmath38 irreducible representations over @xmath27 .
we define a formal power series @xmath42 , called the * hilbert series * of @xmath41 , that records the class of @xmath43)$ ] in the representation ring of @xmath44 for all @xmath30 .
our main result is a rationality result for this series .
the strongest and most general result takes some preparation to state , so we confine ourselves to the following simplified form here ( which is a special case of theorem [ fsghilb ] ) : [ mainthm : fs2 ] let @xmath41 be a finitely generated representation of @xmath5 over an algebraically closed field @xmath27 .
let @xmath45 be the exponent of the group @xmath4 .
then @xmath46 can be written in the form @xmath47 , where @xmath19 and @xmath48 are polynomials in the @xmath49 with coefficients in @xmath50 , and @xmath48 factors over @xmath51 as @xmath52 , where @xmath53 is a @xmath54$]-linear combination of the @xmath49 .
( and @xmath55 is a primitive @xmath45th root of unity . ) to paraphrase : if @xmath41 is a finitely generated @xmath5-module then the representations @xmath43)$ ] of @xmath44 satisfy recursive relations of a very particular form .
we show this by connecting these hilbert series for the generating functions of `` quasi - ordered languages '' which we introduce in this paper . to give a sense of how this connects to our previous paper @xcite : there we showed that finitely generated @xmath56-representations have hilbert series similar to those of @xmath57 , except with @xmath58 in place of @xmath54 $ ] .
the difference is that such hilbert series are controlled by `` ordered languages '' , and quasi - ordered languages are built from these with additional congruence conditions , which is why roots of unity appear .
this is by far the deepest result of the paper , and much of our work goes into its proof .
the idea is to first use brauer s theorem to reduce to the case where @xmath4 is a cyclic group whose order is invertible in @xmath27 .
representations of @xmath4 are identified with vector spaces graded by the character group @xmath59 of @xmath4 , and , in a similar fashion , representations of @xmath5 with those of @xmath60 , a certain category whose objects are finite sets in which each element is weighted by an element of @xmath59 .
an ordered version of this category is amenable to the theory of lingual structures developed in @xcite , which produces rationality results as above for hilbert series .
we did not state any theorems about hilbert series of @xmath17-modules .
when @xmath4 is finite , one can use the fact that the hilbert series of a finitely generated @xmath17-module is also the hilbert series of a finitely generated @xmath0-module ( use proposition [ prop : fig - propertyf ] ) , together with known results on hilbert series of ordinary @xmath0-modules ( * ? ? ?
* theorem b ) , ( * ? ? ?
* corollary 7.1.7 ) .
* the category @xmath61 is equivalent to the category @xmath62 defined in ( * ? ? ?
it is possible to define and prove properties about modified versions of our categories to include her category @xmath63 ; we leave the modifications to the reader .
we point the reader to @xcite for applications of the category @xmath62 .
* applications of @xmath17-modules to topology and group theory are studied in @xcite . * in [ sec : wreath - stab ] , we show that wreath product generalizations of murnaghan s stability theorem follow from the noetherian property for @xmath17-modules . *
in [ ss : twisted - stab ] , we use the machinery developed in @xcite applied to the category @xmath17 to prove general twisted homological stability results for wreath products @xmath64 when @xmath4 is a polycyclic - by - finite group .
* let @xmath41 be a simply - connected manifold of dimension at least @xmath65 . in @xcite
, it is shown that the rational homotopy groups of the configuration spaces of @xmath41 satisfy representation stability , i.e. , are finitely generated @xmath0-modules ( see @xcite ) . in
[ ss : homotopy - conf ] , we outline how this result might be extended when we drop the assumption that @xmath41 is simply - connected by using the category @xmath66 . * as we explain in [ ss : gendelta ] , representations of the category @xmath5 when @xmath4 is a symmetric group are essentially equivalent to @xmath3-modules , in the sense of @xcite .
this observation was our original source of motivation for studying the category @xmath5 : since @xmath5-modules are much easier to think about than @xmath3-modules , this point of view could be profitable . indeed , our results on @xmath5-modules imply the main theorems about @xmath3-modules from @xcite , and more , and are more straightforward than the proofs given there .
these theorems apply for any @xmath4 , though , and so represent a significant generalization of the theory of @xmath3-modules . * the main theorem on hilbert series of @xmath3-modules in @xcite , theorem b , was suspected to be suboptimal .
our original motivation in proving theorem [ mainthm : fs2 ] was to improve ( * ? ? ? * theorem b ) , which it does .
we subsequently found an even stronger improvement , which appears in ( * ? ? ? * theorem 9.2.3 ) .
however , ( * ? ? ?
* theorem 9.2.3 ) is very specific to @xmath3-modules , while theorem [ mainthm : fs2 ] applies to any group @xmath4 .
let @xmath4 be a finite group and let @xmath27 be an algebraically closed field .
an interesting problem is to determine the smallest ( or at least a small ) field @xmath67 with the following property : if @xmath41 is a finitely generated @xmath5-module over @xmath27 then @xmath46 can be written in the form @xmath47 where @xmath68 $ ] and @xmath69 factors as @xmath70 where each @xmath71 is a linear combination of the @xmath72 with coefficients in the ring of integers @xmath73 .
we prove that one can take @xmath74 , where @xmath45 is the exponent of @xmath4 .
in fact , we show that @xmath74 if @xmath4 is `` @xmath45-good . '' for example , @xmath75 is 2-good ( in good characteristic ) , and so @xmath76 in this case . in characteristic 0 , it follows from example [ hilbexample ] that @xmath77 must contain the field generated by the character table of @xmath4 .
can one always take @xmath77 to be the field generated by the brauer character table of @xmath4 ?
theorem [ mainthm : fi3 ] shows that @xmath28 knows very little of @xmath4 ( only the number of irreducible representations ) .
in contrast , @xmath40 knows a lot about @xmath4 : for instance , it knows about tensor products of @xmath4 representations .
it seems reasonable to think one could recover @xmath4 from @xmath78 .
here is a precise question .
let @xmath4 and @xmath79 be finite groups and let @xmath27 be an algebraically closed field .
suppose that @xmath40 and @xmath80 are equivalent as @xmath27-linear abelian categories .
are @xmath4 and @xmath79 isomorphic ? on a first reading
, we encourage the reader to go through [ sec : fig ] and [ sec : fsg ] to see the main results .
[ sec : fig ] is mostly self - contained except for the background material on representations of categories in [ sec : background ] . in [ sec : background ] , we review material that we will use often , especially the main results from @xcite . in [ sec : fig ] we investigate the category @xmath17 , and prove theorems [ mainthm : fi1 ] , [ mainthm : fi2 ] , and [ mainthm : fi3 ] . in
[ ss : qordered ] , we introduce a class of formal languages , the * quasi - ordered languages * , and prove results about their hilbert series . in [ sec : weighted - surj ] , we discuss representations of the category @xmath81 of finite weighted sets .
quasi - ordered languages are used to establish the main result about hilbert series of representations of this category , and this result about hilbert series is the key input to the proof of theorem [ mainthm : fs2 ] . in [ sec : fsg ] , we study @xmath57 .
we prove theorems [ mainthm : fs1 ] and [ mainthm : fs2 ] , discuss the connection to @xmath3-modules , and give some examples .
* @xmath82 denotes the set of non - negative integers . * for @xmath83 ,
define @xmath20 = \{1 , \dots , n\}$ ] with the convention @xmath84 = \emptyset$ ] .
* @xmath85 is the completion of the symmetric algebra , so @xmath86 .
* for a partition @xmath87 , we let @xmath88 denote the specht module , which is a representation of the symmetric group @xmath89 with @xmath90 ( and is irreducible in characteristic @xmath91 )
. see ( * ? ? ?
* chapter 4 ) for details ; there it is denoted @xmath92 $ ] . we thank jeremy miller for explaining the construction in [ ss : homotopy - conf ] .
we also thank aurlien djament , nathalie wahl , and a referee for helpful comments .
in this section , we recall the main combinatorial preliminaries that we need from @xcite . we also prove some additional results .
let @xmath93 be an essentially small category .
we denote by @xmath94 the set of isomorphism classes in @xmath93 .
we say that @xmath93 is * directed * if every self - map in @xmath93 is the identity .
if @xmath93 is directed , then @xmath94 is naturally a poset by defining @xmath95 if there exists a morphism @xmath96 .
fix a nonzero ring @xmath27 ( not necessarily commutative ) and let @xmath97 denote the category of left @xmath27-modules .
a * representation * of @xmath93 over @xmath27 ( or a * @xmath98$]-module * ) is a functor @xmath99 .
a map of @xmath93-modules is a natural transformation .
we write @xmath100 for the category of representations , which is abelian .
let @xmath101 be an object of @xmath93 .
define a representation @xmath102 of @xmath93 by @xmath103 $ ] , i.e. , @xmath104 is the free left @xmath27-module with basis @xmath105 .
if @xmath41 is another representation then @xmath106 .
this shows that @xmath107 is an exact functor , and so @xmath102 is a projective object of @xmath100 .
we call it the * principal projective * at @xmath101 .
a @xmath93-module is * finitely generated * if it is a quotient of a finite direct sum of principal projective objects .
an object of @xmath100 is * noetherian * if every ascending chain of subobjects stabilizes ; this is equivalent to every subrepresentation being finitely generated .
the category @xmath100 is * noetherian * if every finitely generated object in it is .
let @xmath108 be a functor .
this gives a pullback functor on representations @xmath109 defined by @xmath110 .
we are interested in how @xmath111 interacts with finiteness properties of representations . the following definition ( which is ( * ? ? ?
* definition 3.2.1 ) ) is of central importance to this .
[ def : propf ] we say that @xmath112 satisfies * property ( f ) * if the following condition holds : given any object @xmath101 of @xmath93 there exist finitely many objects @xmath113 of @xmath114 and morphisms @xmath115 in @xmath93 such that for any object @xmath116 of @xmath114 and any morphism @xmath117 in @xmath93 , there exists a morphism @xmath118 in @xmath114 such that @xmath119 . the following result shows why property ( f ) is so useful .
the next three statements are ( * ? ? ?
* propositions 3.2.3 , 3.2.4 , corollary 3.2.5 ) : [ propffg ] the functor @xmath112 satisfies property ( f ) if and only if @xmath111 takes finitely generated objects of @xmath100 to finitely generated objects of @xmath120 .
recall that @xmath112 is essentially surjective if every object of @xmath93 is isomorphic to @xmath121 for some object @xmath122 .
[ pullback - fg ] suppose that @xmath112 is essentially surjective .
let @xmath41 be an object of @xmath100 such that @xmath123 is finitely generated ( resp .
noetherian ) . then @xmath41 is finitely generated ( resp .
noetherian ) . [ pullback - noeth ]
suppose that @xmath120 is noetherian and @xmath112 is essentially surjective and satisfies property ( f ) .
then @xmath100 is noetherian .
[ propfcomp ] consider functors @xmath124 and @xmath125 . 1 .
if @xmath126 satisfy property ( f ) , then the composition @xmath127 satisfies property ( f ) .
if @xmath112 is essentially surjective and @xmath127 satisfies property ( f ) , then @xmath128 satisfies property ( f ) .
this follows immediately from the previous three propositions .
a * norm * on @xmath93 is a function @xmath129 , where @xmath130 is a finite set .
a * normed category * is a category equipped with a norm .
fix a category @xmath93 with a norm @xmath131 with values in @xmath132 ; let @xmath133 be indeterminates .
let @xmath41 be a representation of @xmath93 over a field @xmath27 .
we define the * hilbert series * of @xmath41 as @xmath134 when this makes sense .
we omit the norm @xmath131 from the notation when possible .
a poset @xmath135 is * noetherian * if for every sequence @xmath136 of elements in @xmath135 , there exists @xmath137 such that @xmath138 .
see @xcite for basic facts . for an object @xmath101 ,
let @xmath139 be the functor given by @xmath140 .
note that @xmath141 $ ] .
an * ordering * on @xmath142 is a choice of well - order on @xmath143 , for each @xmath144 , such that for every morphism @xmath145 in @xmath93 the induced map @xmath146 is strictly order - preserving .
we write @xmath147 for an ordering ; @xmath142 is * orderable * if it admits an ordering .
set @xmath148 given @xmath149 and @xmath150 , define @xmath151 if there exists @xmath152 with @xmath153 . define an equivalence relation on @xmath154 by @xmath155 if @xmath151 and @xmath156 .
the poset @xmath157 is the quotient of @xmath154 by @xmath158 , with the induced partial order .
[ def : grobner - cat ] we say that @xmath93 is * grbner * if , for all objects @xmath101 , the functor @xmath142 is orderable and the poset @xmath157 is noetherian .
we say that @xmath93 is * quasi - grbner * if there exists a grbner category @xmath114 and an essentially surjective functor @xmath159 satisfying property ( f ) .
the following statement is ( * ? ? ?
* theorem 4.3.2 ) : [ grobnoeth ] let @xmath93 be a quasi - grbner category . then for any left - noetherian ring @xmath27 , the category @xmath100 is noetherian .
the following results follow easily from the definitions .
[ potgrob ] suppose that @xmath108 is an essentially surjective functor satisfying property ( f ) and @xmath114 is quasi - grbner .
then @xmath93 is quasi - grbner .
[ grobprod ] the cartesian product of finitely many ( quasi-)grbner categories is ( quasi-)grbner .
in this section , we assume that @xmath4 is finite . define a functor @xmath160 that sends a set to itself and a function @xmath161 to @xmath162 where @xmath163 is the constant map sending every element to the identity of @xmath4 .
we also use @xmath112 to denote the restriction @xmath164 .
we have the following basic property about representations of @xmath17 : [ prop : fig - propertyf ] the functors @xmath160 and @xmath164 satisfy property ( f ) .
in particular , @xmath16 and @xmath17 are quasi - grbner categories , and if @xmath27 is left - noetherian then @xmath165 and @xmath166 are noetherian . to see that @xmath112 satisfies property ( f ) , suppose @xmath101 has size @xmath167 , set @xmath168 all equal to @xmath101 and let @xmath169 correspond to all automorphisms of @xmath101 in @xmath16 of the form @xmath170 under some enumeration . the categories @xmath0 and @xmath171 are quasi - grbner by ( * ? ? ?
* theorems 7.1.4 , 7.3.4 ) , so the last part follows from proposition [ potgrob ] .
corollary [ cor : polycyclic ] improves the noetherian conclusion by allowing @xmath4 to be any polycyclic - by - finite group .
recall from @xcite that @xmath172 is the category of finite sets where a morphism @xmath173 is an injection @xmath174 and a function @xmath175 with composition defined in the evident way .
define a category @xmath176 of finite sets whose morphisms are pairs @xmath177 where @xmath19 is a decorated injection as in the definition of @xmath172 and @xmath178 is as in the definition of @xmath17 . as above , there is a natural functor @xmath179 satisfying property ( f ) .
the category @xmath28 only depends on @xmath78 as an abelian category equipped with the extra structure of the invariants functor @xmath180 .
thus @xmath28 `` sees '' very little of @xmath4
. we can sometimes be more explicit .
let @xmath39 be the groupoid of finite sets ( maps are bijections ) .
[ figrep ] suppose that @xmath27 is a splitting field for @xmath4 in which @xmath181 is invertible .
then representations of @xmath17 are equivalent to representations of @xmath37 , where @xmath38 is the number of non - trivial irreducible representations of @xmath4 over @xmath27 .
let @xmath182 be the non - trivial irreducible representations of @xmath4 , and let @xmath183 be the trivial representation .
suppose that @xmath41 is an @xmath17-module .
since @xmath27 is a splitting field for @xmath4 , @xmath184 $ ] is isomorphic to a direct product of matrix algebras ; since @xmath185 \cong \bk[g]^{\otimes s}$ ] , we see that @xmath27 is also a splitting field for all @xmath186 .
so we can decompose @xmath187 into isotypic pieces for the action of @xmath186 : @xmath188 where @xmath45 is a multiplicity space .
suppose now that @xmath174 is an injection . to build a morphism in @xmath17
we must also choose a function @xmath189 .
however , if @xmath178 and @xmath190 are two choices then @xmath177 and @xmath191 differ by an element of @xmath192 , namely an automorphism of the form @xmath193 where @xmath194 restricts to @xmath195 on @xmath7 .
thus it suffices to record the action of @xmath196 . note that if @xmath197 restricts to 1 on @xmath7 then @xmath198 .
it follows that @xmath162 must map @xmath187 into the @xmath199-invariants of @xmath200 . in other words , under the above decomposition , @xmath162 induces a linear map @xmath201 thus , associated to @xmath41 we have built a representation @xmath45 of @xmath37 .
the above discussion makes clear that no information is lost in passing from @xmath41 to @xmath45 , and so this is a fully faithful construction .
the inverse construction is defined by the formula . by proposition [ figrep ]
, an @xmath17-module can be thought of as a sequence @xmath202 , where each @xmath203 is an @xmath0-module equipped with an action of @xmath204 .
there are no transition maps , so in a finitely generated @xmath17-module , all but finitely many of the @xmath203 are zero .
thus , at least in good characteristic , @xmath17-modules are not much different from @xmath0-modules , and essentially any result about @xmath0-modules ( e.g. , noetherianity ) carries over to @xmath17-modules .
there are some similarities between proposition [ figrep ] and the results of @xcite .
we need one last thing for our application in the next section .
suppose that @xmath27 is commutative .
given @xmath17-modules @xmath41 and @xmath45 , their * pointwise tensor product * , denoted @xmath205 is the @xmath17-module given by @xmath206 and @xmath207 for morphisms @xmath19 .
[ prop : segre - fig ] for any commutative ring @xmath27 , the pointwise tensor product of two finitely generated @xmath17-modules is finitely generated . by (
* proposition 3.3.2 ) , it suffices to prove that the diagonal functor @xmath208 satisfies property ( f ) . to see this ,
let @xmath101 and @xmath209 be finite sets and consider @xmath4-injections @xmath210 let @xmath211 and define a morphism @xmath212 where @xmath213 is the inclusion and @xmath214 denotes the constant map @xmath215 . also define @xmath216 and @xmath217 to be the maps induced by @xmath19 and @xmath218 . then @xmath219 and @xmath220 .
since @xmath221 and @xmath4 is finite , there are only finitely many choices for @xmath222 , @xmath223 , and @xmath224 up to isomorphism and taking all of these choices shows that @xmath3 satisfies property ( f ) .
using proposition [ figrep ] and following @xcite , we can deduce a generalization of murnaghan s stability theorem for tensor products of representations of symmetric groups to arbitrary wreath products . for this section ,
set @xmath225 , and assume that @xmath4 is finite
. for basics on complex representations of wreath products of finite groups , see ( * ? ? ?
* chapter 5 ) .
let @xmath226 be the set of isomorphism classes of irreducible representations of @xmath4 .
the irreducible representations @xmath227 of the wreath product @xmath64 are indexed by partition - valued functions @xmath228 on @xmath226 such that @xmath229 .
[ rmk : wr - translate ] to relate this with proposition [ figrep ] , we make the following observations .
first , given a representation @xmath230 of @xmath4 and a representation @xmath231 of @xmath89 , the vector space @xmath232 carries an action of @xmath64 : for @xmath233 , @xmath234 , @xmath235 , and @xmath236 , we have @xmath237 this extends to an action of @xmath64 on @xmath232 ; write this representation as @xmath238 .
the representation @xmath227 can be constructed as follows . for each @xmath239 , let @xmath240 ; @xmath241 is a representation of @xmath242 , and this group can be identified with a subgroup of @xmath64 .
the induction is @xmath227 .
if we think of @xmath227 as a @xmath17-module concentrated in degree @xmath167 , then we see from the description just given that , under the bijection in proposition [ figrep ] , @xmath227 corresponds to the tensor product of the @xmath243 , thought of as an @xmath37-module . given a partition @xmath87 and a positive integer @xmath167 , define @xmath244 = ( n-|\lambda| , \lambda_1 , \lambda_2 , \dots)$ ] ( we assume that @xmath167 is large enough so that this is a partition ) . given any partition - valued function @xmath228 on @xmath226 , let @xmath245 $ ] denote the function which agrees with @xmath228 on all non - trivial representations and is @xmath246 $ ] on the trivial representation @xmath247 .
[ thm : wr - stab ] fix partition - valued functions @xmath248 on @xmath226 .
the multiplicity of @xmath249)$ ] in @xmath250 ) \otimes v(\ul{\mu}[n])$ ] as representations of @xmath251 is independent of @xmath167 for @xmath252 .
we first recall a few facts about @xmath0-modules from @xcite ( that paper is not written in the language of @xmath0-modules , but see ( * ? ? ?
* proposition 1.3.5 ) ) .
let @xmath253 be the category of finitely generated torsion @xmath0-modules ( for finitely generated modules , torsion is equivalent to being nonzero only in finitely many degrees ) and let @xmath254 be the category of finitely generated @xmath0-modules .
then every object in @xmath255 has finite length ( * ? ? ?
* corollary 2.2.6 ) .
furthermore , the pointwise tensor product of two finitely generated @xmath0-modules is also finitely generated ( * ? ? ?
* proposition 2.3.6 ) , and hence the pointwise tensor product of two finite length objects in @xmath256 still has finite length . for every partition @xmath87 , there is an @xmath0-module @xmath257 generated in degree @xmath258 whose value on a set of size @xmath167 is the representation @xmath259}$ ] ( see @xcite ) .
let @xmath260 be the image of @xmath257 in @xmath256 ; @xmath260 is a simple object , and all simple objects are of this form ( * ? ? ?
* corollary 2.2.7 ) .
a representation @xmath261 is naturally a functor on @xmath39 and functors on @xmath37 can be built by taking the tensor product of functors on @xmath0 and @xmath38 functors on @xmath39 .
for every partition - valued function @xmath228 on @xmath226 , define a functor @xmath262 on @xmath37 which is the tensor product of @xmath263 on @xmath0 and @xmath264 on the copy of @xmath39 which is indexed by @xmath230 . from remark [ rmk : wr - translate ]
, using the equivalence in proposition [ figrep ] , this corresponds to a functor on @xmath17 whose value on a set of size @xmath167 is @xmath250)$ ] if @xmath167 is large enough for @xmath245 $ ] to be a partition - valued function ( and @xmath91 otherwise ) .
write @xmath265 . using the facts for @xmath0 and @xmath256 recalled above ,
every object of @xmath266 has finite length and the simple objects are of the form @xmath267 .
furthermore , the pointwise tensor product preserves finite generation in @xmath17 ( proposition [ prop : segre - fig ] ) , so @xmath268 has a finite filtration by modules of the form @xmath269 in @xmath266 .
finally , we note that the modules in @xmath270 are concentrated in finitely many degrees .
so the filtration encodes the usual tensor product for large enough degree .
this implies the desired stability result .
when @xmath4 is trivial , the proof of theorem [ thm : wr - stab ] generalizes the one given in @xcite and when @xmath271 , theorem [ thm : wr - stab ] is proven in ( * ? ? ?
* theorem 5.3 ) . in this section
we prove theorem [ mainthm : fi1 ] .
define @xmath272 to be the category whose objects are finite ordered sets and where morphisms are pairs @xmath8 as in @xmath17 , except that @xmath19 is required to be order - preserving . we first prove : [ oig : noeth ] suppose that @xmath29 $ ] is left - noetherian for all @xmath30 .
then @xmath273 is noetherian .
it suffices to show that the principal projective @xmath274 of @xmath272 at @xmath20 $ ] is noetherian . as a @xmath27-module
, we have @xmath275)=\bigoplus_{f \colon [ n ] \to [ m],\ a \in g^n } \bk e_{(f , a)}\ ] ] where @xmath276 runs over @xmath277 , [ m])$ ] and @xmath278 is a basis vector . set @xmath279 $ ] .
we give @xmath280)$ ] the structure of a left @xmath6-module by @xmath281 .
let @xmath282 .
then the vectors @xmath283 s form a basis for @xmath280)$ ] as a left @xmath6-module .
suppose that @xmath284 \to [ \ell]$ ] is a morphism in @xmath285 and @xmath286 .
then @xmath287 thus the @xmath285-module structure on @xmath274 is compatible with the left @xmath6-module structure .
now , @xmath280)$ ] also has a left action of @xmath288)=g^m$ ] . if @xmath289 then @xmath290 where @xmath291 is the pullback map .
we can thus write @xmath292 for @xmath293 . in general , @xmath294$]-submodules of @xmath274 need not be @xmath295$]-submodules ( see remark [ rmk : emulate ] ) , but this is true for `` monomial '' subrepresentations , which is enough to prove the theorem , as we now explain .
following @xcite , we order the set of order - preserving injections @xmath20 \to [ m]$ ] ( @xmath167 fixed , @xmath296 varying ) using the lexicographic order . given @xmath297)$ ] where @xmath298 is non - zero and @xmath299,[m])$ ] , with the @xmath300 distinct , define @xmath301 to be @xmath302 where @xmath300 is lexicographically maximal among the @xmath19 s occurring in the sum .
let @xmath41 be a @xmath294$]-submodule of @xmath274 .
define the initial submodule @xmath303 of @xmath41 by setting @xmath304 to be the @xmath27-span of @xmath305 .
by @xcite , @xmath303 is a @xmath306$]-submodule of @xmath274 .
we claim that @xmath303 is also an @xmath6-submodule . indeed , suppose that @xmath307)$ ] .
let @xmath308 be an element of @xmath309)$ ] with @xmath310 .
we assume that @xmath311 , and that the @xmath300 are distinct .
pick @xmath312 .
the forgetful map @xmath313 is surjective , so choose @xmath314 with @xmath315 .
we have @xmath316 since @xmath317 is a unit of @xmath6 , it follows that @xmath318 , and so @xmath319 . we have @xmath320)$ ] , since it is obtained from @xmath101 using the @xmath272-module structure , and so we see that @xmath321 belongs to @xmath322)$ ] , proving the claim .
now suppose that @xmath323 is an ascending chain of @xmath294$]-submodules of @xmath274 .
then @xmath324 is an ascending chain of @xmath295$]-submodules of @xmath274 , and therefore stabilizes since @xmath325 is noetherian ( * ? ? ?
* theorem 7.1.1 ) .
it follows from a standard grbner basis argument ( see ( * ? ? ?
* proposition 4.2.2 ) ) that the chain @xmath326 stabilizes , and so @xmath274 is noetherian .
[ rmk : emulate ] if @xmath41 is a @xmath294$]-submodule of @xmath274 then it need not be closed under the action of @xmath6 by pre - composition : that is , there is no way to emulate pre - composition with automorphisms of @xmath20 $ ] using post - composition with other morphisms .
we do not prove this ( as it is unnecessary to do so ) , but give an illustrative example .
the group @xmath280)$ ] has the structure of a @xmath327$]-module , since @xmath328)$ ] . if @xmath329 $ ] and @xmath330 is in @xmath280)$ ] , with @xmath331 $ ] , then @xmath332 , where @xmath333 \to \bk[g^n]$ ] is induced by @xmath291 .
given @xmath334 , it is not generally possible to pick @xmath329 $ ] such that @xmath335 for all @xmath19 , and so one can not emulate the @xmath6-structure using the @xmath327$]-structure . before proving theorem [ mainthm : fi1 ] , we require two lemmas .
[ oigf ] the forgetful functor @xmath336 satisfies property ( f ) .
let @xmath101 be an object of @xmath17 and @xmath116 an object of @xmath272 .
choose a total order on @xmath101 , and write @xmath337 for the corresponding object of @xmath272 , so that @xmath338 . consider a morphism @xmath339 in @xmath17 .
we can factor @xmath340 as @xmath341 , where @xmath178 is a permutation and @xmath342 is order - preserving .
let @xmath343 , so that @xmath48 defines a morphism @xmath344 in @xmath272 .
let @xmath345 , where @xmath214 is the constant function @xmath346 , so that @xmath347 is a morphism in @xmath17 .
then @xmath348 .
since there are only finitely many choices for @xmath349 , the result follows .
[ twisted - noeth ] let @xmath6 be a ring and let @xmath79 be a group acting on @xmath6 .
suppose that the twisted group algebra @xmath350 $ ] is left - noetherian
. then @xmath6 is left - noetherian .
recall that the twisted group algebra is the set of @xmath6-linear combinations of elements in @xmath79 with the multiplication @xmath351 .
let @xmath41 be a left @xmath6-module .
for @xmath352 , let @xmath353 be the left @xmath6-module with underlying abelian group @xmath41 on which @xmath6 acts by @xmath354 .
let @xmath355 .
then @xmath356 is naturally a left @xmath350$]-module .
suppose now that @xmath357 is an ascending chain of left ideals of @xmath6 .
then @xmath358 is an ascending chain of left ideals of @xmath350 $ ] , and therefore stabilizes .
this clearly implies that the original chain stabilizes as well , and so @xmath6 is left - noetherian .
first suppose that @xmath29 $ ] is left - noetherian for all @xmath30 .
then @xmath273 is noetherian by theorem [ oig : noeth ] .
thus by lemma [ oigf ] and proposition [ pullback - noeth ] we find that @xmath28 is noetherian .
conversely , suppose that @xmath28 is noetherian .
let @xmath359 $ ] and let @xmath360 \cong \bk[s_n \ltimes g^n]$ ] be the twisted group algebra ( @xmath89 acting on @xmath44 by permutations ) .
then a left @xmath360$]-module is the same as a @xmath361$]-module supported at @xmath167 .
it follows that @xmath360 $ ] is left - noetherian , and so @xmath6 is left - noetherian by lemma [ twisted - noeth ] .
let @xmath4 be a group and @xmath27 be a commutative ring and let @xmath362 be an @xmath17-module .
let @xmath363)$ ] .
consider the morphism @xmath20 \to [ n+1]$ ] consisting of the injection defined by @xmath364 and the trivial @xmath4-map @xmath20\to g$ ] sending every element to the identity .
this gives a @xmath365$]-equivariant map @xmath366 .
[ thm : wr - twisted - stab ] assume that @xmath4 is a polycyclic - by - finite group and that @xmath27 is noetherian , and let @xmath41 be a finitely generated @xmath17-module .
then for each @xmath367 , the map @xmath368 is an isomorphism for @xmath252 .
we apply ( * ? ? ?
* theorem 4.2 ) .
first , @xmath17 is a complemented category where the monoidal structure is given by disjoint union of sets , and the generator is the one element set .
the two conditions of that theorem are : ( 1 ) the category of @xmath17-modules is noetherian , and ( 2 ) the isomorphism above holds for @xmath252 when @xmath41 is the trivial @xmath17-module .
the first point is corollary [ cor : polycyclic ] and the second point is ( * ? ? ?
* proposition 1.6 ) .
define a shift functor @xmath369 on @xmath17-modules by @xmath370 . in (
* definition 4.10 ) , the following definition is introduced ( to avoid ambiguity , we use the phrase @xmath369-degree instead of just degree ) .
first , a functor which takes the value @xmath91 except at finitely many isomorphism classes of objects has @xmath369-degree @xmath371 and , in general , @xmath77 has @xmath369-degree @xmath372 if the kernel and cokernel of the natural map @xmath373 have @xmath369-degree @xmath374 .
then ( * ? ? ?
* theorem 4.20 ) proves theorem [ thm : wr - twisted - stab ] for any @xmath4 under the assumption that @xmath41 has finite @xmath369-degree ( and gives bounds for when stability occurs in terms of the @xmath369-degree ) .
so it is natural to ask if a finitely generated @xmath17-module necessarily has finite @xmath369-degree .
we do not know if this is true , but will now prove it when @xmath4 is finite .
[ prop : finiteg - polynomial ] if @xmath4 is finite , then a finitely generated @xmath17-module has finite @xmath369-degree .
let @xmath41 be an @xmath17-module .
the kernel of @xmath375 is torsion , i.e. , all non - invertible morphisms act by zero .
if @xmath41 is finitely generated , then the same is true for the kernel ( proposition [ prop : fig - propertyf ] ) , and hence it is supported in finitely many degrees .
it is easy to see that a torsion module concentrated in degrees @xmath376 has @xmath369-degree @xmath376 , so from now on , we only need to focus on the cokernel of @xmath375 , which we denote @xmath377 .
note that @xmath3 is right - exact .
also , @xmath378 \otimes_{\bk } p_{s'}\\ \delta p_s & = \bigoplus_{s ' \subset s } \bk[g ] \otimes_{\bk } p_{s'},\end{aligned}\ ] ] where @xmath379 in both sums
. since @xmath4 is finite , we observe that if @xmath41 is finitely generated in degree @xmath376 , then @xmath377 is finitely generated in degree @xmath380 ( here we use that @xmath4 is finite ) . by induction , we are done .
a. proposition [ prop : finiteg - polynomial ] and the discussion above shows that theorem [ thm : wr - twisted - stab ] follows from ( * ? ? ?
* theorem 4.20 ) when @xmath4 is finite .
when @xmath4 is trivial , we have been informed that theorem [ thm : wr - twisted - stab ] is contained in work of betley @xcite and church @xcite also with bounds for when stability occurs . by proposition [ prop : fig - propertyf ] , when @xmath4 is finite , the result also follows from this work .
our result is new when @xmath4 is infinite .
b. using ( * ? ? ?
* theorem b ) , the proof of proposition [ prop : finiteg - polynomial ] can be adapted to work for any group @xmath4 if we replace finitely generated by the condition that it is presented in finite degree . in particular , if @xmath4 is polycyclic - by - finite , the statement of proposition [ prop : finiteg - polynomial ] remains valid .
along the same lines , by @xcite , the result also holds for any @xmath4 in the special case when the @xmath17-module can be extended to an @xmath381-module ( @xmath381 is a larger category where morphisms are partially defined @xmath4-injections between finite sets ) . let @xmath382 be the category of sets with a free @xmath4-action with finitely many orbits and injective @xmath4-equivariant maps .
there is an equivalence @xmath383 defined by @xmath384 ( the @xmath4-action is @xmath385 ) and sending @xmath386 to the morphism @xmath387 given by @xmath388 .
let @xmath41 be a connected manifold with @xmath389 .
fix @xmath390 and set @xmath391 .
let @xmath356 be the universal cover of @xmath41 , so @xmath4 acts on @xmath356 by deck transformations .
given a set @xmath135 with a free @xmath4-action , let @xmath392 be the space of injective @xmath4-equivariant maps .
also , for any set @xmath7 , let @xmath393 be the space of injective maps .
there is a natural map @xmath394 with fiber @xmath395 .
since @xmath389 , @xmath396 and hence @xmath392 is simply - connected .
given an equivariant injective map of sets @xmath397 with free @xmath4-action , we have a forgetful map @xmath398 and hence a map on homotopy groups @xmath399 ( we have not chosen basepoints : each @xmath400 is simply - connected , so there is a canonical isomorphism between the homotopy groups for any two choices of basepoint ) .
so we can define a functor @xmath401 by @xmath402 . using the equivalence above ,
this is also a functor on @xmath403 .
now let @xmath404 be any abelian group .
we define an @xmath17-module by @xmath405 .
when @xmath391 is trivial and @xmath406 , we get an @xmath0-module defined over @xmath407 , which is studied in @xcite .
in particular , they show that these are finitely generated @xmath0-modules .
they prove a similar result when @xmath408 and also for the functor @xmath409 .
it would be interesting to find more general conditions on @xmath41 , @xmath404 , and @xmath410 under which the above @xmath17-module is finitely generated .
let @xmath369 be an * alphabet * ( i.e. , a finite set ) and let @xmath411 be the free monoid generated by @xmath369 .
* words * are elements of @xmath411 and * languages * are subsets of @xmath411 .
we write words as @xmath412 where @xmath413 ; the length of the word is @xmath167 .
a subword of @xmath414 is a word of the form @xmath415 where @xmath416 ( @xmath417 is allowed ) .
we can also represent a word as a function @xmath20 \to \sigma$ ] given by @xmath418 ; equivalently , we can write them as functions @xmath419 where @xmath130 is ordered , this is convenient for defining subwords by restricting the function to @xmath420 $ ] .
let @xmath59 be a finite abelian group and let @xmath421 be a monoid homomorphism .
given a subset @xmath422 , define @xmath423 we say that a language @xmath424 is a * congruence language * if it is of the form @xmath425 for some @xmath59 , @xmath426 and @xmath7 .
the * modulus * of a congruence language is the exponent of the group @xmath59 .
( the * exponent * of a group is the least common multiple of the orders of all of its elements . )
let @xmath427 be a formal power series in variables @xmath428 .
an * @xmath45-cyclotomic translate * of @xmath77 is a series of the form @xmath429 , where @xmath430 are @xmath45th roots of unity .
[ lem : cyc - trans ] let @xmath59 be a finite abelian group of exponent @xmath45 , let @xmath7 be a subset of @xmath59 , and let @xmath431 be a monoid homomorphism .
suppose that @xmath432 is a formal power series over @xmath433 .
let @xmath434 , where the sum is over @xmath435 such that @xmath436 .
then @xmath437 is a @xmath50-linear combination of @xmath45-cyclotomic translates of @xmath427 .
we have @xmath438 , where @xmath439 is the characteristic function of @xmath7 .
now express @xmath440 as a @xmath50-linear combination of characters of @xmath59 . a * norm * on @xmath369 is a monoid homomorphism @xmath441 for some set @xmath130 induced by a function @xmath442 ( i.e. , if @xmath443 then @xmath444 is a standard basis vector ) .
it is * universal * if @xmath442 is injective .
a norm @xmath131 on a language @xmath445 is the restriction of a norm on @xmath369 to @xmath446 .
[ hilb - cong - intersect ] let @xmath446 be a language on @xmath369 equipped with a universal norm @xmath131 with values in @xmath132 , let @xmath447 be a congruence language on @xmath369 of modulus @xmath45 , and let @xmath448 .
then @xmath449 is a @xmath50-linear combination of @xmath45-cyclotomic translates of @xmath450 .
choose @xmath451 and @xmath452 so that @xmath453 .
since @xmath131 is universal , the map @xmath454 can be factored as @xmath455 , where @xmath456 is a monoid homomorphism .
thus if @xmath457 , then @xmath458 is obtained by simply discarding the terms for which @xmath459 .
the result now follows from lemma [ lem : cyc - trans ] .
recall from @xcite that an * ordered language * on @xmath369 is a language obtained from the singleton languages and the languages @xmath460 for @xmath461 , using finite union and concatenation . a * quasi - ordered language * ( of modulus @xmath45 )
is the intersection of an ordered language and a congruence language ( of modulus @xmath45 ) ; we say that the language is @xmath462 .
the class of quasi - ordered languages is not closed under unions , intersections , or concatenations , but here is a positive result ( we will not use regular languages , so see @xcite for the definition ) : 1 .
congruence languages are regular .
quasi - ordered languages are regular .
\(a ) pick an isomorphism @xmath463 .
then @xmath426 can be factored as @xmath464^\star \to \lambda$ ] where @xmath465 $ ] denotes the set of generators for @xmath59 under the above isomorphism .
the inverse image of @xmath7 under @xmath465^\star \to \lambda$ ] consists of all words where the @xmath38-tuple of the multiplicities of each @xmath466 $ ] taken modulo @xmath467 is one of finitely many values .
it is clear that this can be encoded by a dfa with @xmath468 many states ( see @xcite ) , so is a regular language .
hence @xmath469 is the inverse image of a regular language in @xmath465^\star$ ] under a monoid homomorphism , and thus is regular ( * ? ? ?
* theorem 3.5 ) .
\(b ) ordered languages are regular by ( * ? ? ?
* theorem 5.3.1 ) .
now use the fact that the intersection of regular languages is regular .
[ defn : peck ] let @xmath470 be an integer .
we say that @xmath471 is of class @xmath472 if it can be written as @xmath47 where @xmath19 and @xmath48 are polynomials in the @xmath49 with coefficients in @xmath50 and @xmath48 factors as @xmath473 , where @xmath71 is a @xmath54$]-linear combination of the @xmath49 .
our main result on quasi - ordered languages is the following theorem .
[ qordhilb ] let @xmath446 be a quasi - ordered language of modulus @xmath45 equipped with a norm valued in @xmath132 .
then @xmath474 is of class @xmath472 .
this follows immediately from ( * ? ? ?
* theorem 5.3.7 ) and proposition [ hilb - cong - intersect ] . for later applications
, it will be convenient to formulate a coordinate - free version of @xmath472 .
suppose that @xmath475 is a finite rank free @xmath58-module and @xmath476 .
we say that @xmath19 is @xmath472 if there is a @xmath58-basis @xmath477 of @xmath475 so that @xmath19 is @xmath472 as a series in the @xmath49 .
this is independent of the choice of @xmath58-basis of @xmath475 .
suppose that @xmath93 is directed and normed over @xmath132 and pick an object @xmath101 of @xmath93 .
we define a norm on @xmath157 as follows : given @xmath478 , let @xmath479 be a lift , and put @xmath480 .
this is well - defined because @xmath93 is directed : if @xmath481 is a second lift , then @xmath116 and @xmath482 are isomorphic .
a * @xmath462-lingual structure * on @xmath157 is a pair @xmath483 consisting of a finite alphabet @xmath369 normed over @xmath132 and an injection @xmath484 compatible with the norms , i.e. , such that @xmath485 and such that for every poset ideal @xmath486 of @xmath157 , the language @xmath487 is @xmath462 . following ( * ? ? ?
* definition 6.3.1 ) , a category @xmath93 is * @xmath462-lingual * if @xmath488 admits a @xmath462-lingual structure for all objects @xmath489 .
[ hilbthm ] let @xmath93 be a @xmath490-lingual grbner category and let @xmath41 be a finitely generated representation of @xmath93 .
then @xmath46 is @xmath472 , i.e. , is a rational function @xmath47 , where @xmath491 and @xmath69 are polynomials with coefficients in @xmath50 and @xmath69 factors as @xmath492 and each @xmath493 is a @xmath54$]-linear combination of the @xmath49 .
this follows directly from ( * ? ? ?
* theorem 6.3.2 ) and theorem [ qordhilb ] . a normed category @xmath93 is * strongly qo@xmath494-lingual * if for each object @xmath101 there exists a @xmath462-lingual structure @xmath484 and a congruence language @xmath447 on @xmath369 such that for every poset ideal @xmath486 of @xmath157 , the language @xmath487 is the intersection of an ordered language with @xmath447 .
( if we drop the adjective `` strongly , '' then the congruence language @xmath447 is allowed to depend on the ideal @xmath486 . )
we have the following variant of ( * ? ? ?
* proposition 6.3.3 ) : [ prop : prod - qon ] let @xmath495 and @xmath496 be strongly @xmath490-lingual normed categories .
suppose that the posets @xmath497 and @xmath498 are noetherian for all @xmath101 and @xmath116 .
then @xmath499 is strongly @xmath490-lingual .
let @xmath500 be an object of @xmath501 , and let @xmath502 be a strongly @xmath462-lingual structure at @xmath500 .
let @xmath503 , normed over @xmath132 in the obvious manner .
let @xmath504 and @xmath505 be the given congruence languages of modulus @xmath45 on @xmath506 and @xmath507 , regarded as languages on @xmath369 . write @xmath508 , where @xmath509 .
let @xmath510 be the map defined by @xmath511 for @xmath512 , and @xmath513 for @xmath514 .
let @xmath515 , and let @xmath516 .
then @xmath447 is a congruence language on @xmath369 of modulus @xmath45 and has the following property : if @xmath517 and @xmath518 are any languages on @xmath506 and @xmath507 , then @xmath519 .
the following diagram commutes : @xmath520 & \vert \cc_{1,{x_1 } } \vert \times \vert \cc_{2,x_2 } \vert \ar[r]^-{i_1 \times i_2 } \ar[d ] & \sigma_1^{\star } \times \sigma_2^{\star } \ar[r ] \ar[d ] & \sigma^{\star } \ar[d ] \\ & \bn^{i_1 } \oplus \bn^{i_2 } \ar@{=}[r ] & \bn^{i_1 } \oplus \bn^{i_2 } \ar@{=}[r ] & \bn^i } \ ] ] the top right map is concatenation of words .
we let @xmath521 be the composition along the first line , which is clearly injective . we claim that this is a strongly @xmath462-lingual structure on @xmath522 .
the commutativity of the above diagram shows that it is a lingual structure .
now suppose @xmath486 is an ideal of @xmath522 .
since this poset is noetherian ( being a direct product of noetherian posets ) , @xmath486 is a finite union of principal ideals @xmath523 .
each @xmath524 is of the form @xmath525 , where @xmath526 is an ideal of @xmath527 and @xmath528 is an ideal of @xmath529 . by assumption , @xmath530 is the intersection of an ordered language with @xmath504 , while @xmath531 is the intersection of an ordered language with @xmath505 .
it follows from ( * ? ? ?
* theorem 5.3.1 ) that the direct product of two ordered languages on disjoint alphabets is also ordered , so @xmath532 is the intersection of an ordered language with @xmath447 . by definition ,
finite unions of ordered languages are ordered , thus @xmath487 is the intersection of an ordered language with @xmath447 .
in this section we study a generalization of the category of finite sets with surjective functions by considering weighted sets .
this is preparatory material for the next section .
let @xmath59 be a finite abelian group .
a * weighting * on a set @xmath7 is a function @xmath534 .
a * weighted set * is a set equipped with a weighting ; we write @xmath535 to denote the weighting if we need to distinguish between sets .
suppose that @xmath426 is a weighting on @xmath7 , and let @xmath174 be a function .
we define @xmath536 to be the weighting on @xmath537 given by @xmath538 .
a * map of weighted sets * @xmath173 is a _ surjective _ function @xmath174 such that @xmath539 . in particular , if there is a morphism @xmath173 , then @xmath540 .
we let @xmath81 denote the category of finite weighted sets .
we require an ordered version of the category as well .
let @xmath533 be the following category .
the objects are finite ordered weighted sets .
the order and weighting are not required to interact in any way .
a morphism @xmath173 is a surjective function @xmath174 such that @xmath539 , and for all @xmath541 in @xmath537 we have @xmath542 .
we define a norm @xmath131 on @xmath533 taking values in @xmath543 by defining @xmath544 to be the function @xmath545 ( thought of as an element of @xmath543 ) .
our main result about @xmath533 is : [ thm : ows ] the category @xmath546 is grbner and strongly @xmath490-lingual , where @xmath45 is the exponent of @xmath59 .
we prove this in the next section , and now use it to study @xmath81 .
[ thm : fws - qg ] the category @xmath60 is quasi - grbner . the forgetful functor @xmath547 is easily seen to satisfy property ( f ) : given @xmath548 , @xmath549 , and @xmath550 , there is a unique ordering on @xmath7 so that @xmath19 is order - preserving , so @xmath551 where the sum is over all choices @xmath552 of orderings on @xmath7 .
so the result follows from theorem [ thm : ows ] .
[ fws - noeth ] if @xmath27 is left - noetherian then @xmath553 is noetherian .
we now give a generalization of the above theorem that we will need in our analysis of @xmath5 .
let @xmath41 be a finitely generated @xmath554-module .
enumerate @xmath555 as @xmath556 , and let @xmath557 be a formal variable corresponding to @xmath558
. given @xmath559 , let @xmath560 $ ] be the @xmath555-weighted set where @xmath561 elements have weight @xmath558 .
when @xmath27 is a field , define the hilbert series of @xmath41 by @xmath562 , \ldots , [ \bn(r ) ] ) \cdot \bt_1^{\bn(1 ) } \cdots \bt_r^{\bn(r)},\ ] ] where for @xmath563 we write @xmath564 for the multinomial coefficient @xmath565 the reason for the multinomial coefficient is in the proof of the next theorem , which is the main result we need in our applications .
[ fswhilb ] let @xmath41 be a finitely generated @xmath554-module and assume @xmath27 is a field .
then @xmath46 is @xmath472 ( see definition [ defn : peck ] ) where @xmath45 is the exponent of @xmath566 .
using theorem [ thm : ows ] , @xmath567 is grbner by proposition [ grobprod ] , and is ( strongly ) @xmath462-lingual by proposition [ prop : prod - qon ] .
now the result follows from theorem [ thm : ows ] using a version of the functor @xmath112 from the proof of theorem [ thm : fws - qg ] .
there are @xmath564 isomorphism classes in @xmath568 which map to the isomorphism class of @xmath560 $ ] in @xmath569 ; thus , with the multinomial coefficients , we have @xmath570 . fix a finite set @xmath571 and let @xmath572 .
given @xmath573 and @xmath574 , we write @xmath575 for the corresponding element of @xmath369 .
we denote elements of @xmath411 by @xmath576 , where @xmath577 and @xmath578 are words of equal length . for @xmath573 , we define @xmath579 by @xmath580 we let @xmath581 be @xmath582 .
note that @xmath583 and @xmath584 are monoid homomorphisms . for @xmath585 ,
we let @xmath586 be the set of all @xmath587 with @xmath588 .
this is a congruence language of modulus @xmath45 ( the exponent of @xmath59 ) .
we now define a partial order on @xmath589 .
let @xmath590 \to \sigma$ ] and @xmath591 \to \sigma$ ] be two words .
define @xmath592 if there exists an ordered surjection @xmath593 \to [ n]$ ] ( i.e. , @xmath137 implies @xmath594 for all @xmath595 $ ] ) such that @xmath596 and @xmath597 , i.e. , @xmath598 for @xmath599 $ ] ; we say @xmath19 * witnesses * @xmath592 .
note that if @xmath592 then @xmath600 .
let @xmath601 be a word in @xmath589 .
we say that an index @xmath602 is _ special _ for @xmath101 if @xmath603 for any @xmath604 , that is , the letter @xmath605 appears nowhere before @xmath602 in @xmath606 .
[ lem : sum-1 ] let @xmath95 be two words in @xmath589 of lengths @xmath167 and @xmath296 .
suppose that the final letters of @xmath101 and @xmath116 are equal , say to @xmath575 .
suppose furthermore that @xmath167 is special for @xmath101 if and only if @xmath296 is special for @xmath116 .
then we can find an ordered surjection @xmath593 \to [ n]$ ] witnessing @xmath95 and satisfying @xmath607 .
write @xmath601 and @xmath608 , and let @xmath284 \to [ n]$ ] be an ordered surjection witnessing @xmath95 , so that @xmath609 and @xmath610 . if @xmath317 appears only once in @xmath101 and @xmath116 then @xmath611 , so we can take @xmath612 .
suppose now that @xmath317 appears at least twice in both @xmath101 and @xmath116 .
we consider two cases .
first suppose that @xmath613 .
write @xmath614 for some @xmath615 $ ] .
we note that @xmath616 .
let @xmath617 be such that @xmath618 .
define @xmath593 \to [ n]$ ] by @xmath619 for @xmath620 and @xmath621 for @xmath622 .
then @xmath607 .
it is clear that @xmath19 is a surjection and satisfies @xmath623 .
we have @xmath624 for all
@xmath625 , and so @xmath19 is ordered since @xmath48 is .
finally , @xmath626 since @xmath194 sums to 0 over @xmath7 . thus @xmath19 witnesses @xmath95 . now suppose that @xmath627 .
write @xmath628 and @xmath629 .
since @xmath630 , it follows that @xmath7 is non - empty and @xmath631 .
we have @xmath632 and @xmath633 .
since @xmath634 by assumption , we see that @xmath635 .
define @xmath593 \to [ n]$ ] by @xmath619 for @xmath636 , @xmath621 for @xmath637 , and @xmath638 .
then @xmath607 . we have already explained that @xmath626 , and so the reasoning of the previous paragraph shows that @xmath19 is an ordered surjection witnessing @xmath95 .
[ lem : sum-2 ] let @xmath95 be two words in @xmath589 of lengths @xmath167 and @xmath296 , and let @xmath639 .
suppose that : 1 .
the words formed from the final @xmath38 letters of @xmath101 and @xmath116 are equal .
2 . for @xmath640 ,
the index @xmath641 is special for @xmath101 if and only if @xmath642 is for @xmath116 .
then we can find an ordered surjection @xmath593 \to [ n]$ ] witnessing @xmath95 and satisfying @xmath643 for @xmath640 .
we prove the lemma by induction on @xmath38 .
the @xmath644 case is trivial .
suppose now that @xmath645 and that the lemma has been proven for all smaller values of @xmath38 . by lemma [ lem : sum-1 ] , we can find an ordered surjection @xmath646 \to [ n]$ ] witnessing @xmath95 and satisfying @xmath647 .
let @xmath209 and @xmath648 be the words formed by deleting the final letters of @xmath101 and @xmath116
. then @xmath649 , as witnessed by @xmath650}$ ] .
by the inductive hypothesis , we can find @xmath651 \to [ n-1]$ ] witnessing @xmath649 and satisfying @xmath652 for @xmath653 .
we now define @xmath593 \to [ n]$ ] by @xmath654 for @xmath655 $ ] and @xmath638 .
if @xmath101 is a word of length @xmath167 then we write @xmath656 for @xmath657 .
for instance , @xmath658 is the rightmost letter in @xmath101 .
[ lem : sum-4 ] let @xmath101 and @xmath116 be two words in @xmath589 of lengths @xmath167 and @xmath296 , let @xmath659 be an integer , and let @xmath660 .
suppose that conditions ( a ) and ( b ) from lemma [ lem : sum-2 ] hold .
suppose furthermore that @xmath649 , where @xmath209 is obtained from @xmath101 by deleting indices @xmath661 , and @xmath648 is formed similarly from @xmath116
. then @xmath95 .
we first handle the case @xmath662 , where it suffices to treat the case @xmath663 , and we assume this to ease notation .
the inequality @xmath649 is witnessed by some ordered surjection @xmath664 \setminus \{m - r+1\ } \to [ n ] \setminus \{n - r+1\}$ ] . by lemma [ lem : sum-2 ]
, we can assume that @xmath665 for @xmath666 .
we now define @xmath593 \to [ n]$ ] by @xmath654 for @xmath667 and @xmath668 .
this is an ordered surjection witnessing @xmath95 .
now we do @xmath669 general by induction on @xmath669 .
let @xmath670 be obtained from @xmath101 be deleting letters @xmath671 , and let @xmath672 be obtained from @xmath116 similarly
. then @xmath209 is obtained from @xmath670 by deleting index @xmath673 ( using the indexing in @xmath101 ) , and @xmath648 is obtained similarly from @xmath672 .
note that @xmath673 is special in @xmath670 if and only if it is in @xmath101 , since the two words only differ to the right of @xmath673 .
it follows that the specialness of @xmath673 is the same for @xmath670 and @xmath672 .
the @xmath662 case shows that @xmath674 . by induction on @xmath669
we now have @xmath95 .
[ lem : sum-0 ] any @xmath578 with @xmath675 contains nonempty subsequences that sum to @xmath91 .
the partial sums @xmath676 for @xmath677 must contain a repeated element , and the difference of these two partial sums is the desired set of elements that sums to @xmath91 .
[ lem : sum-5 ] let @xmath601 be a word in @xmath589 , and put @xmath678 .
then we can find @xmath679 such that @xmath680 and @xmath681 .
some letter of @xmath571 , say @xmath317 , must appear at least @xmath682 times in the final @xmath38 letters of @xmath606 .
suppose that @xmath683 are such that @xmath684 for all @xmath685 .
put @xmath686 . by lemma [ lem : sum-0 ]
, we can take @xmath687 to be some subset of @xmath688 .
[ prop : sigmastar - noeth ] the poset @xmath589 is noetherian . we apply nash williams theory @xcite . recall that an infinite sequence @xmath689 is _ bad _ if @xmath690 whenever @xmath691 , and _ minimal bad _ if it is bad and for each @xmath167 there is no bad sequence @xmath692 with @xmath693 .
suppose that @xmath589 is not noetherian .
then a minimal bad sequence @xmath689 exists : this is a standard consequence of @xmath589 being well - founded ( no infinite decreasing sequences ) , which is obvious .
let @xmath694 be the set of words @xmath116 such that @xmath695 for some @xmath602 .
then @xmath696 is noetherian : if not , pick a bad sequence @xmath697 in @xmath7 .
then @xmath698 for some @xmath699 and then @xmath700 is a bad sequence in @xmath411 which violates minimality of @xmath701 .
we now proceed to reach a contradiction .
let @xmath702 .
let @xmath703 be a subsequence of @xmath689 such that @xmath704 for all @xmath602 , the final @xmath38 letters of @xmath705 are independent of @xmath602 , and the specialness of the final @xmath38 indices is independent of @xmath602 .
such a subsequence exists since @xmath706 and there are only finitely many possibilities for the final @xmath38 letters and their specialness .
write @xmath707 .
fix @xmath708 such that @xmath709 and @xmath710 for all @xmath602 .
such numbers exist by lemma [ lem : sum-5 ] and the fact that the final @xmath38 letters of @xmath705 are independent of @xmath602 .
let @xmath711 be obtained from @xmath705 by deleting the letters at @xmath661 .
then @xmath712 for all @xmath602 .
indeed , suppose @xmath713 , and put @xmath714 .
we regard @xmath711 as indexed by @xmath20 \setminus t$ ] .
the ordered surjection @xmath715 \to [ n_i ] \setminus t$ ] defined by @xmath716 for @xmath717 and @xmath718 for @xmath637 then witnesses the claimed inequality .
the main point here is that @xmath719 , which ensures the appropriate identity on @xmath720 . since @xmath721 for all
@xmath602 and @xmath7 is noetherian , there exists @xmath691 such that @xmath722 .
but then @xmath723 by lemma [ lem : sum-4 ] , which contradicts the badness of the original sequence @xmath689 .
it follows that @xmath589 is noetherian .
fix @xmath585 .
let @xmath724 be a word in @xmath586 ( recall this means that @xmath725 ) .
put @xmath726 where @xmath727 means any element of @xmath59 .
define a language @xmath728 by @xmath729 it is clear that @xmath728 is an ordered language .
[ lem : ows-2 ] if @xmath730 then @xmath592 .
let @xmath730 be a word of length @xmath296 and write @xmath731 , with @xmath732 .
let @xmath733 $ ] be the indices occurring in the words @xmath734 and let @xmath735 be the complement of @xmath486 , so that @xmath736 .
define a function @xmath593 \to [ n]$ ] as follows .
on @xmath735 , we let @xmath19 be the unique order - preserving bijection . for @xmath737 ,
let @xmath738 $ ] be minimal so that @xmath739 . now define @xmath19 on @xmath486 by @xmath740 .
it is clear that @xmath19 is an ordered surjection and that @xmath623 .
since @xmath741 , it follows that @xmath742 . from the way we defined @xmath19
, it follows that @xmath743 .
thus @xmath626 , which completes the proof .
we say that a word @xmath744 is * minimal * if no non - empty subsequence of @xmath745 sums to @xmath91 .
note that we started with the second index . by convention ,
words of length @xmath214 are minimal . by lemma [ lem : sum-0 ] , there are only finitely many minimal words .
let @xmath576 in @xmath586 be given .
we say that @xmath591 \to \sigma^{\star}$ ] is * minimal * over @xmath590 \to \sigma^{\star}$ ] if there is an ordered surjection @xmath593 \to [ n]$ ] such that @xmath596 and @xmath597 and for every @xmath599 $ ] the word @xmath746 is minimal .
if @xmath747 is minimal over @xmath576 then the length of @xmath747 is bounded , so there are only finitely many such minimal words .
[ lem : ows-3 ] let @xmath748 be words in @xmath586 .
there exists @xmath747 minimal over @xmath576 such that @xmath749 .
let @xmath750 be the lengths of @xmath576 and @xmath751 , and choose a witness @xmath593 \to [ n]$ ] to @xmath748 .
let @xmath752 $ ] be the set of elements of the form @xmath753 for @xmath599 $ ] .
let @xmath754 $ ] be minimal subject to @xmath755 and @xmath756 .
then @xmath757 is minimal for all @xmath599 $ ] . indeed ,
if it were not then we could discard a subsequence summing to @xmath91 and make @xmath735 smaller .
we thus see that @xmath758 is minimal over @xmath576 . if @xmath759 \setminus k$ ] then there exists @xmath604 in @xmath130 with @xmath760 , and so @xmath749 . [ lem : ows-4 ] every poset ideal of @xmath586 is of the form @xmath761 , where @xmath446 is an ordered language on @xmath369 .
it suffices to treat the case of a principal ideal .
thus consider the ideal @xmath7 generated by @xmath762 .
let @xmath763 for @xmath764 be the words minimal over @xmath576 , and let @xmath765 .
then @xmath446 is an ordered language , by construction . if @xmath766 then @xmath767 for some @xmath602 , and so @xmath768 by lemma [ lem : ows-2 ] , and so @xmath769 .
conversely , suppose @xmath769 .
then @xmath770 for some @xmath602 by lemma [ lem : ows-3 ] , and of course @xmath771 , and so @xmath766 .
the category @xmath772 is clearly directed .
let @xmath773 , \theta)$ ] be an object of @xmath93 . to show that @xmath93 is grbner , we need to show that @xmath774 is orderable and is a noetherian poset .
lexicographic order on @xmath589 induces an admissible order on @xmath774 . to show that @xmath774 is noetherian , we apply the above theory with @xmath775 $ ] .
suppose that @xmath161 is a map in @xmath93 , with @xmath776 , \phi)$ ] ; note that this means that @xmath19 is a surjection @xmath777 \to [ n]$ ] .
we define a word @xmath777 \to \sigma^{\star}$ ] by mapping @xmath778 $ ] to @xmath779 .
one can reconstruct @xmath19 from this word , so this defines an injection @xmath780 .
in fact , the image lands in @xmath586 .
it is clear from the definition of the order on @xmath589 that @xmath602 is strictly order - preserving .
thus @xmath774 is noetherian by proposition [ prop : sigmastar - noeth ] .
finally , since @xmath602 maps ideals to ideals , we see that it gives a strong qo@xmath494-lingual structure on @xmath774 by lemma [ lem : ows-4 ] .
in this section we study the surjective version of @xmath4-sets . the noetherian property is deduced in [ ss : fsg - basic ] . in [ ss : gendelta ] , we explain the connection between @xmath5-modules and @xmath3-modules . in [ ss : fsexample ] , we give some examples of @xmath5-modules .
the results on hilbert series are stated in [ sec : gsurj - results ] .
the remainder of the section is devoted to proving the hilbert series results .
let @xmath785 be given .
say that a morphism @xmath786 is minimal if the function @xmath787 is injective .
since @xmath4 is finite , minimal implies @xmath788 , so there are finitely many minimal maps up to isomorphism .
now consider a map @xmath786 in @xmath18 .
define an equivalence relation on @xmath116 by @xmath789 if @xmath790 and @xmath791 , and let @xmath792 be the quotient .
then the induced map @xmath793 is minimal .
furthermore , @xmath794 . reversing all of the morphisms , we see that @xmath795 satisfies property ( f ) .
let @xmath800 be an abelian category equipped with a symmetric `` cotensor '' structure , i.e. , a functor @xmath801 , and analogous data opposite to that of a tensor structure .
( here we are using the deligne tensor product of abelian categories @xcite . )
given a surjection @xmath802 of finite sets , there is an induced functor @xmath803 by cotensoring along the fibers of @xmath19 .
a * @xmath3-module * over @xmath800 is a rule @xmath41 that assigns to each finite set @xmath7 an object @xmath804 of @xmath805 and to each surjection @xmath802 of finite sets a morphism @xmath806 , such that if @xmath802 and @xmath807 are surjections , then the diagram @xmath808^{m_g } \\ ( gf)^*(m_r ) \ar[ur]^{f^*(m_g ) } \ar[rr]^{m_{gf } } & & m_t } \ ] ] commutes .
there are two main examples relevant to this paper : * let @xmath800 be the category of polynomial functors @xmath809 .
then @xmath810 is identified with the category of polynomial functors @xmath811 .
there is a comultiplication @xmath812 taking a functor @xmath77 to the functor @xmath813 , and this gives @xmath800 the structure of a symmetric cotensor category .
@xmath3-modules over @xmath800 are @xmath3-modules as defined in @xcite .
* let @xmath800 be the category of representations of a finite group @xmath4 .
then @xmath810 is identified with the category of representations of @xmath814 .
there is a comultiplication @xmath812 taking a representation @xmath230 of @xmath4 to the representation @xmath815 of @xmath814 , where @xmath4 is included in @xmath814 via the diagonal map .
this gives @xmath800 the structure of a symmetric cotensor category .
@xmath3-modules over @xmath800 are representations of @xmath5 .
if @xmath816 is invertible in the base field then the category of polynomial functors of degree @xmath817 is equivalent , as a cotensor category , to the category @xmath818 . to see this
, first note that invertibility of @xmath816 implies that @xmath819 for @xmath820 , as well as the category of polynomial functors of degree @xmath820 are semisimple categories , and are equivalent by schur
weyl duality ( this is usually stated over characteristic @xmath91 @xcite , but all that is needed is @xmath816 to be invertible ) .
the compatibility of the cotensor structures follows from the form of the schur
weyl equivalence .
we thus find that @xmath3-modules of degree @xmath817 ( in the sense of @xcite ) coincide with representations of @xmath821 .
thus our results on @xmath5 can be loosely viewed as a generalization of our results on @xmath3-modules from @xcite ( `` loosely '' because in bad characteristic the results are independent of each other ) .
it seems possible that our results could generalize to @xmath3-modules over any `` finite '' abelian cotensor category .
we now give a source of @xmath5-modules .
let @xmath135 and @xmath822 be simplicial complexes on finite vertex sets @xmath823 and @xmath824 . define a simplicial complex @xmath825 on the vertex set @xmath826 as follows .
let @xmath827 be the projection map , and similarly define @xmath828 .
then @xmath829 is a simplex if and only if @xmath830 and @xmath831 are simplices of @xmath135 and @xmath822 and have the same cardinality as @xmath7 .
we call @xmath825 the * segre product * of @xmath135 and @xmath822
. it is functorial for maps of simplicial complexes .
it is not a topological construction , and depends in an essential way on the simplicial structure .
fix a finite simplicial complex @xmath135 , equipped with an action of a group @xmath4 .
the diagonal map @xmath832 induces a map of simplicial complexes @xmath833 .
we thus obtain a functor from @xmath5 to the category of simplicial complexes by @xmath834 .
fixing @xmath602 , we obtain a representation @xmath835 of @xmath5 by @xmath836 .
it is not difficult to directly show that @xmath837 is a finitely generated representation of @xmath5 , where @xmath838 denotes the space of simplicial @xmath602-chains .
thus by corollary [ fsg - noeth ] , @xmath835 is a finitely generated representation of @xmath5 .
theorem [ fsghilb ] below gives information about the hilbert series of @xmath835 .
the case where @xmath135 is just a single simplex is already extremely complicated and interesting , and is closely related to syzygies of the segre embedding .
in fact , if @xmath135 has @xmath839 vertices then the @xmath840-module given by @xmath841 coincides with the degree @xmath839 piece of the @xmath3-module @xmath842 of @xmath669-syzygies of the segre embedding ( as defined in @xcite ) under the equivalence in the previous section , at least when @xmath843 is invertible in @xmath27 .
let @xmath41 be a finitely generated @xmath798-module over an algebraically closed field @xmath27 .
we need the following fact : given finite groups @xmath4 and @xmath79 , every irreducible @xmath844$]-module is of the form @xmath845 where @xmath230 and @xmath231 are irreducible modules for @xmath184 $ ] , respectively @xmath846 $ ] ( * ? ? ? * proposition 2.3.23 ) .
this implies that we have a canonical identification @xmath847 let @xmath848 , and write @xmath560 $ ] for @xmath849)_{i \in i}$ ] .
then @xmath850})$ ] is a finite - dimensional representation of @xmath851 .
let @xmath852_{\bn}$ ] denote the image of the class of this representation under the map @xmath853 where @xmath854 note that in good characteristic , one can recover the isomorphism class of @xmath855)$ ] as a representation of @xmath851 from @xmath852_{\bn}$ ] due to the @xmath856-equivariance . if @xmath857 are the irreducible representations of the @xmath858 , then @xmath852_{\bn}$ ] can be thought of as a polynomial in corresponding variables @xmath859 .
define the * hilbert series * of @xmath41 by @xmath860_{\bn}.\ ] ] this is an element of the ring @xmath861 .
this definition does not fit into our framework of hilbert series of normed categories , though it can be seen as an enhanced hilbert series ( similar to @xcite ) .
[ fsghilb ] let @xmath41 be a finitely generated representation of @xmath798 over an algebraically closed field @xmath27 .
then @xmath46 is a @xmath472 function of the @xmath72 , where @xmath45 is the least common multiple of the exponents of the @xmath858 not divisible by @xmath669 . stating the full result requires some additional notions .
let @xmath4 be a finite group and let @xmath27 be a field .
let @xmath862 be a collection of subgroups such that the orders of the commutator subgroups @xmath863 $ ] are invertible in @xmath27 , and let @xmath864 $ ] be the abelianization of @xmath865 .
there is a functor @xmath866 given by taking coinvariants under @xmath867 $ ] .
this functor is exact since the order of @xmath863 $ ] is invertible in @xmath27 , and thus induces a homomorphism of grothendieck groups @xmath868 .
there are also homomorphisms @xmath869 given by restriction .
we say that the family @xmath870 is * good * if the composite @xmath871 is a split injection ( i.e. , an injection with torsion - free cokernel ) .
we say that @xmath4 is * @xmath45-good * if it admits a good family @xmath870 such that the exponent of each @xmath872 divides @xmath45 .
we say that a family @xmath873 of finite groups is @xmath45-good if each member is .
[ fsghilb2 ] suppose that @xmath873 is @xmath45-good .
let @xmath41 be a finitely generated @xmath798-module over an algebraically closed field @xmath27 .
then @xmath46 is a @xmath472 function of the @xmath557 . using brauer s theorem ,
we show that over an algebraically closed field , every group is @xmath45-good where @xmath45 is prime - to-@xmath669 part of the exponent of the group ( proposition [ fsglem2 ] ) , and so theorem [ fsghilb ] follows from theorem [ fsghilb2 ] .
we show that symmetric groups are 2-good if @xmath816 is invertible in @xmath27 , which recovers some of our results on hilbert series of @xmath3-modules ( see @xcite ) in good characteristic . for general groups , we know little about the optimal value of @xmath45 .
finding some results could be an interesting group theory problem .
[ hilbexample ] let @xmath4 be a finite group and let @xmath874 be the set of irreducible representations of @xmath4 over @xmath433 .
define an @xmath5-module @xmath835 by @xmath875 where @xmath876 is the diagonal map .
let @xmath877 be the set of conjugacy classes in @xmath4 , @xmath878 be the character of @xmath879 , and @xmath49 be an indeterminate corresponding to @xmath879 .
a computation similar to that in ( * ? ? ?
5.7 ) gives @xmath880 this is a @xmath472 function of the @xmath49 , as predicted by theorem [ fsghilb ] , where @xmath45 is such that all characters of @xmath4 take values in @xmath50 .
let @xmath881 . if @xmath882 then , by convention , every group has order prime to @xmath669 , and the only @xmath669-group is the trivial group .
we say that a collection @xmath883 of subgroups of @xmath4 is a * covering * if the map on grothendieck groups @xmath884 is a split injection .
recall that if @xmath885 is a prime , then an * @xmath885-elementary group * is one that is the direct product of an @xmath885-group and a cyclic group of order prime to @xmath885 .
an * elementary group * is a group which is @xmath885-elementary for some prime @xmath885 . 1 .
let @xmath883 be a covering of @xmath4 , and for each @xmath602 let @xmath886 be a covering of @xmath887 , and let @xmath888 .
then @xmath889 is a covering of @xmath4 .
2 . let @xmath883 be a covering of @xmath4 , and suppose each @xmath887 is @xmath45-good .
then @xmath4 is @xmath45-good .
3 . suppose that @xmath79 is a @xmath669-elementary group and write @xmath890 , where @xmath891 is cyclic of order prime to @xmath669 and @xmath892 is a @xmath669-group .
then @xmath893 is a covering of @xmath79 . 1 .
the collection of elementary subgroups @xmath883 of @xmath4 is a covering of @xmath4 .
if @xmath79 has order prime to @xmath669 , then @xmath79 is @xmath45-good where @xmath45 is the exponent of @xmath79 .
\(d ) let @xmath896 be the restriction map .
let @xmath897 be the grothendieck group of finite - dimensional projective @xmath184$]-modules .
the map @xmath898 given by @xmath899 is a perfect pairing @xcite . combining this with frobenius reciprocity
, it follows that the dual of @xmath900 can be identified with the induction map @xmath901 .
this map is surjective by brauer s theorem ( * ? ? ?
* , thm .
since the dual of @xmath900 is surjective , it follows that @xmath900 is a split injection , which proves the claim .
\(e ) indeed , arguing with duals and frobenius reciprocity again , it is enough to find subgroups @xmath902 of @xmath79 such that the induction map @xmath903 is surjective .
( note that @xmath904 for groups of order prime to @xmath669 . )
this follows from brauer s theorem ( * ? ? ?
* , thm .
20 ) . by parts ( a ) , ( c ) , and ( d ) of lemma [ lem : good - groups ]
, @xmath4 has a covering by its subgroups of order prime to @xmath669 . for each of these groups , its set of subgroups is good by part ( e ) .
now finish by applying ( b ) .
we now construct a good collection of subgroups for the symmetric group @xmath89 in good characteristic . given a partition @xmath905 with @xmath906 ,
let @xmath907 be the corresponding young subgroup of @xmath89 . under the assumption on @xmath910 , the representations of @xmath89 are semisimple .
using frobenius reciprocity , the restriction map on representation rings is dual to induction .
we claim that each irreducible character of @xmath89 is a @xmath58-linear combination of the permutation representations on @xmath911 ( this implies @xmath908 is a good collection of subgroups ) . recall that the irreducible representations of @xmath89 are indexed by partitions of @xmath167 ( we denote them @xmath88 ) .
also , recall the dominance order on partitions : @xmath912 if @xmath913 for all @xmath602 . an immediate consequence of pieri s rule ( * ? ? ?
* ( 2.10 ) ) is that the permutation representation @xmath911 contains @xmath88 with multiplicity @xmath214 and the remaining representations @xmath261 that appear satisfy @xmath914 .
this proves the claim .
first , we use the good family of subgroups to reduce to the case where each @xmath858 is abelian of invertible order . for such @xmath4 ,
we identify @xmath916-modules with @xmath917-modules , where @xmath59 is the group of characters of @xmath4 .
the theorem then follows from our results for hilbert series of @xmath917-modules .
we now go through the details .
[ fsglem3 ] let @xmath796 be finite groups . for each @xmath602 ,
let @xmath887 be a subgroup of @xmath858 such that the order of @xmath918 $ ] is invertible in @xmath27 .
define a functor @xmath919 by letting @xmath920 be the @xmath921^{\ul{s}}$]-coinvariants of @xmath922 .
then we have the following : 1 . @xmath923 is a well - defined object of @xmath924 .
if @xmath41 is finitely generated then so is @xmath923 .
3 . let @xmath925 be the map induced by restricting to @xmath887 followed by taking @xmath918$]-coinvariants , and let @xmath926 be the sum of the @xmath927 .
then @xmath928 is the image of @xmath929 under the ring homomorphism @xmath930 induced by @xmath426 .
\(a ) for a tuple @xmath931 of sets , let @xmath932 be the @xmath27-subspace of @xmath922 spanned by elements of the form @xmath933 with @xmath934^{\ul{s}}$ ] and @xmath935 .
if @xmath936 is a morphism in @xmath937 then the induced map @xmath938 carries @xmath933 to @xmath939 .
thus @xmath735 is a @xmath940-submodule of @xmath41 , and so @xmath941 is a well - defined @xmath940-module .
the group actions clearly carry through , and so @xmath923 is well - defined .
\(b ) suppose that @xmath942 is finitely generated .
then the restriction of @xmath41 to @xmath940 is finitely generated by propositions [ fsgpropg ] and [ propffg ] .
the restriction of @xmath923 to @xmath940 is a quotient of the restriction of @xmath41 , and is therefore finitely generated .
thus @xmath923 is finitely generated , by proposition [ pullback - fg ] .
pick a finite set @xmath7 .
let @xmath944 be a surjection .
then this can be factored as @xmath945 where the first map is the image of a morphism @xmath946 under @xmath947 .
so for the @xmath948 in the definition of property ( f ) , we take the pairs @xmath949 of subsets of @xmath7 whose union is all of @xmath7 .
[ fsglem4 ] let @xmath796 be groups , let @xmath950 be a surjection , and let @xmath951 be the resulting family of groups indexed by @xmath486 .
let @xmath952 be the functor induced by disjoint union , i.e. , @xmath953 where @xmath954 . 1 .
@xmath112 satisfies property ( f ) ; in particular , if @xmath41 is finitely generated then so is @xmath123 .
2 . let @xmath955 be the diagonal map , and let @xmath956 be the sum of the @xmath927 .
then @xmath957 is the image of @xmath929 under the ring homomorphism @xmath958 induced by @xmath426 .
consider the commutative diagram of categories @xmath959^{\phi } & { \mathbf{fs}}^{\op}_{\ul{g } } \\ ( { \mathbf{fs}}^{\op})^j \ar[r]^{\phi ' } \ar[u ] & ( { \mathbf{fs}}^{\op})^i \ar[u ] } .\ ] ] the functor @xmath960 is defined just like @xmath112 ; it satisfies property ( f ) by proposition [ fsuniong ] .
the vertical maps satisfy property ( f ) by proposition [ fsgpropg ] .
thus @xmath112 satisfies property ( f ) by proposition [ propfcomp ] .
this proves ( a ) ; ( b ) is clear .
[ fsglem5 ] suppose that @xmath796 is a family of commutative groups of exponents dividing @xmath45 .
suppose that @xmath45 is invertible in @xmath27 and that @xmath27 contains the @xmath45th roots of unity .
let @xmath961 be the group of characters of @xmath858 , and let @xmath962 .
then there is an equivalence @xmath963 respecting hilbert series , i.e. , @xmath964 for @xmath965 . before giving the proof , we offer two clarifications .
first , @xmath966 denotes the category @xmath967 .
an object of this category is a tuple of sets @xmath931 equipped with a weight function @xmath968 for each @xmath602 .
second , @xmath929 and @xmath928 are both series in variables indexed by the characters of the @xmath858 .
this is why they are comparable .
let @xmath41 be a representation of @xmath798 .
let @xmath931 be a tuple of sets .
then we have a decomposition @xmath969 where the sum is over weightings @xmath970 of @xmath971 , and @xmath972 is the subspace of @xmath973 on which @xmath974 acts through @xmath970 . if @xmath975 is a morphism in @xmath798 then the map @xmath976 carries @xmath972 into @xmath977 .
we define @xmath923 to be the functor on @xmath978 which assigns to a weighted set @xmath979 the space @xmath972 .
this construction can be reversed : given a representation @xmath41 of @xmath978 , we can build a representation of @xmath798 by defining @xmath973 to be the sum of the @xmath972 .
we leave to the reader the verification that these constructions are quasi - inverse to each other .
this shows that @xmath112 is an equivalence .
it is clear that it preserves hilbert series : we note that the multinomial coefficients in the definition of @xmath928 count , for each @xmath972 , the number of @xmath980 where @xmath981 is a permutation of @xmath970 .
let @xmath41 be a finitely generated representation of @xmath986 , where @xmath796 . for each @xmath987
, let @xmath988 be a good collection of subgroups of @xmath873 such that the exponent of each @xmath872 divides @xmath45 .
let @xmath989 and let @xmath950 be the projection map .
then we have functors @xmath990 let @xmath991 be the image of @xmath41 under the composition . by lemmas [ fsglem3 ] and [ fsglem4 ] , @xmath991 is finitely generated and @xmath992 is the image of @xmath929 under the ring homomorphism corresponding to the natural additive map @xmath993 . by lemma [ fsglem5 ] and
theorem [ fswhilb ] , @xmath992 is @xmath472 .
thus by lemma [ fsglem1 ] , @xmath929 is @xmath472 .
this completes the proof . | we study representations of wreath product analogues of categories of finite sets .
this includes the category of finite sets and injections ( studied by church , ellenberg , and farb ) and the opposite of the category of finite sets and surjections ( studied by the authors in previous work ) .
we prove noetherian properties for the injective version when the group in question is polycyclic - by - finite and use it to deduce general twisted homological stability results for such wreath products and indicate some applications to representation stability .
we introduce a new class of formal languages ( quasi - ordered languages ) and use them to deduce strong rationality properties of hilbert series of representations for the surjective version when the group is finite . |
recent advances in experimental investigations of hadron structure arouse the interest in the theoretical study of the hadron electromagnetic form factors at high momentum transfers ( see e.g. review @xcite and references therein ) . in this connection
the jlab program of investigations on elastic electron - deuteron scattering experiments at @xmath0 10 ( gev / c)@xmath1 ( @xmath2 is transferred momentum ) @xcite attracts considerable attention .
there exists a hope that these jlab experiments will help to determine the limits of application for the two - nucleon model and to clarify the interplay between nucleon - nucleon and quark approaches to the deuteron . using the asymptotic expansion presented in this paper
we show @xcite that the momentum transfer region in the jlab experiments is asymptotical for the deuteron considered as a nucleon nucleon system .
that is why the study of the electromagnetic deuteron form factors is interesting at @xmath3 .
the present work is devoted to the theoretical investigation of deuteron form factors asymptotic behavior at high momentum transfer in the framework of the two - nucleon model .
the form factors asymptotics is studied in accordance with the next points .
a rule , calculation of the form factors asymptotic behavior in the relativistic approaches reduces to the asymptotic estimation of @xmath4-tuple integrals . in the relativistic approach used in our work
the deuteron form factors are expressed in terms of double integrals where integrands peak at the integration domain bound , and corresponding point is not a point of extremum . in this connection
the theorem defining asymptotic expansion of @xmath4-tuple integrals with such integrand is proven in our paper . \2 .
in general , high momentum transfers require relativistic consideration
. however we begin the consideration of the asymptotic estimation of electromagnetic deuteron form factors with the nonrelativistic case and nonrelativistic impulse approximation at @xmath3 .
this is due to the facts that , at first , the nonrelativistic calculation is a less complicated and , second , this calculation is important for establishment of the role of the relativistic effects .
the asymptotic expansion of the relativistic deuteron form factors is calculated in the relativistic invariant impulse approximation in a variant of instant form of poincare - invariant quantum mechanics ( piqm ) developed in our papers previously @xcite .
the relativistic calculations are performed by analogy with nonrelativistic case .
it is shown that relativistic effects essentially slow down the asymptotical decrease of the form factors .
it is found that obtained in the framework of the two - nucleon model relativistic asymptotics coincides with the experimental data .
this paper is organized as follows .
section 2 is devoted to the proving of the central for this work theorem defining asymptotics of multiple integrals of some special type . in sec .
3 a brief review of the formulas for the deuteron form factors in the nonrelativistic and relativistic invariant impulse approximation is given .
the deuteron form factors asymptotics is calculated in nonrelativistic and relativistic impulse approximation with the help of the proven theorem in sec .
4 . in sec .
5 asymptotics of the form factors is obtained for the deuteron wave functions in the conventional representation as a discrete superposition of yukawa - type terms @xcite .
6 contains the conclusions of this paper .
in the following we will consider integrals of the kind : @xmath5dx\ ; , \label{intlap}\ ] ] where @xmath6 is a domain in * r*@xmath7 , @xmath8 , @xmath9 is a large positive parameter .
we will use following definitions : @xmath10 is a bound of the domain @xmath6 , @xmath11=\omega\cup\partial\omega$ ] , the bound @xmath12 if in the vicinity of any point @xmath13 it can be specified by equation @xmath14 , @xmath15 is a neighborhood of a point @xmath16 , and the function @xmath17 in @xmath15 .
let us prove now a lemma and a theorem on the asymptotic estimation of integrals in eq.([intlap ] ) .
l e m m a. _ let @xmath18 be a smooth function in the @xmath11 $ ] , @xmath19 be a continuous function in the @xmath11 $ ] , and @xmath20 , @xmath21}s(\lambda , x)<\infty\;,\ ] ] at some @xmath22 the integral _ ( [ intlap ] ) _ be absolutely convergent : @xmath23dx<\infty\;,\ ] ] and the following conditions be fulfilled at @xmath24 : @xmath25 _ @xmath26 then at @xmath24 the following estimation is valid : @xmath27 p r o o f. at @xmath24 the following estimations are true : @xmath28 @xmath29 from conditions ( [ uslovie1]),([uslovie2 ] ) we obtain the inequality : @xmath30 ) of the lemma is proved .
later we will consider function @xmath18 described in lemma which has the maximal value in the point @xmath13 , and @xmath31 in the vicinity of @xmath32 .
this point is not the point of extremum , that means the validity of the following conditions : @xmath33 and matrix of coefficients @xmath34 : @xmath35 gives the negative determined quadratic form . in eqs .
( [ cond1 ] ) , ( [ cond2 ] ) @xmath36 specifies the internal normal derivative @xmath37 to the @xmath38 , and @xmath39 is an orthonormal basis in the tangential to the @xmath10 plane @xmath40 at the @xmath32 point . for convenience
let us choose in the vicinity of point @xmath32 a frame @xmath41 , so that @xmath32 is the origin of coordinate and the internal normal to @xmath10 coincides with the last basis vector of the new coordinate system .
functions @xmath42 in these coordinates we denote as @xmath43 , and @xmath44 is an image of @xmath45 ( that is an image of a half - vicinity of the point @xmath32 ) .
the equation for @xmath46 in the vicinity of the point @xmath47 can be written in the following way : @xmath48 with @xmath15 is a vicinity of the point @xmath49 , @xmath50 and at @xmath51 .
t h e o r e m. _ let the following conditions be fulfilled : _
1@xmath52).$ ] 2@xmath53 @xmath54 _ is maximal value of function _
@xmath55 @xmath56 , _ and _ @xmath32 _ is not point of extremum . _
3@xmath57 _ in the vicinity of the point _
@xmath32 .
4@xmath53 _ the taylor expansion of functions @xmath58 and @xmath59 in the vicinity of point @xmath32 satisfy the following relations : _
@xmath60\ ; , \label{f*ly}\ ] ] @xmath61 _ the matrix @xmath62 angle brackets denote bilinear form : @xmath63_. _ then at _ @xmath64 _ the following asymptotic expansion is valid : _ @xmath65\sum\limits_{k=0}^{\infty}a_k(\lambda)\;. \label{ocenka}\ ] ] _ the way to calculate coefficients @xmath66 will be determined later . _
p r o o f. let us divide the integral ( [ intlap ] ) into two integrals .
integration domain of the first one is the half - vicinity @xmath45 of the point @xmath32 , and integration domain of the second one is a remainder of integration domain of the original integral .
it is easily shown by the proven lemma that the second integral is exponentially small as compared with @xmath67 $ ] .
so we will estimate asymptotically the first integral only . in the expansion of the function @xmath68 in line with
the condition ( [ s*ly ] ) linear components are absent , because the point @xmath49 is a point of maximum of the function @xmath68 in the region @xmath15 .
let us choose @xmath45 in accordance with inequalities @xmath69 at @xmath70 .
then we can represent the integral ( [ intlap ] ) within exponentially decreasing terms : @xmath71dy\;. \label{idok}\ ] ] let us rewrite integral in eq .
( [ idok ] ) in the following way : @xmath72 with @xmath73f^*(y)dy_n\;. \label{intfi}\ ] ] the integral ( [ intfi ] ) is one - dimensional , and the function @xmath74 reaches extremum on the boundary @xmath75 .
asymptotic expansion of this integral can be found through integration by parts .
after @xmath76 integration we obtain the sequence : @xmath77 \left.\exp[s^*(\lambda , y)]\right|_{\varphi(y')}^{\delta}-\int\limits_{\varphi(y')}^{\delta } { m^n\left[\frac{f^*(\lambda , y)}{s^{*'}(\lambda , y)}\right]'}\exp[s^*(\lambda , y)]dy_n\;,\ ] ] with @xmath78 is a unit operator and @xmath79 the substitution of @xmath75 provides the main contribution to the asymptotics , the value of @xmath80 is exponentially small as compared with the previous .
further integration under these conditions gives the following expansion for the function ( [ intfi ] ) : @xmath81\sum\limits_{k=0}^\infty m^k\left[\frac{f^*(\lambda , y',\varphi(y'))}{s^{*'}(\lambda , y',\varphi(y'))}\right]\;.\ ] ] so @xmath82 m^k\left[\frac{f^*(\lambda , y',\varphi(y'))}{s^{*'}(\lambda , y',\varphi(y'))}\right]dy ' \label{f}\ ] ] the point @xmath49 is an internal point of maximum of the integrand in the expression ( [ f ] ) . functions @xmath68 and @xmath83 satisfy the conditions of lemma ( [ uslovie1 ] ) and theorem ( [ f*ly ] ) , ( [ s*ly ] ) , therefore we can apply a formula for asymptotic estimation of the n - tuple laplas integrals @xcite and obtain corresponding asymptotic expansion ( [ ocenka ] ) of the integral in eq .
( [ f ] ) .
thus the theorem is proven .
note , that in the general case it is rather difficult to write a compact formula for the coefficients @xmath66 in eq .
( [ ocenka ] ) .
they can be obtained such kind of way in any specific cases . as an example
, these coefficients will be obtained and written out explicitly for double integrals in the consideration of the asymptotic estimation of electromagnetic deuteron form factors in sec .
we write out only the first asymptotic term from eq .
( [ ocenka ] ) in the @xmath84 variables : @xmath85\left(\frac{\partial s(\lambda , x^0)}{\partial n}\right)^{-1}|\det b|^{-\frac{1}{2}}f(\lambda , x^0)\ ; , \label{gch}\ ] ] where @xmath86 are defined by conditions ( [ cond1 ] ) and ( [ cond2 ] ) .
in the nonrelativistic impulse approximation known formulas for electromagnetic deuteron form factors can be rewritten in the following way @xcite : @xmath87 @xmath88 @xmath89 here @xmath90 are the deuteron wave functions in momentum representation , @xmath91 0,2 are orbital angular momenta , @xmath92 are nonrelativistic free two - particles charge , quadrupole and magnetic dipole form factors , @xmath93 is the deuteron mass .
formulas for @xmath94 are given in @xcite .
let us discuss briefly possible types of the model deuteron wave functions .
there are several classes of the deuteron wave functions : obtained with microscopic model hamiltonians of the @xmath95-interaction in the non - relativistic nuclear physics ( for example , see @xcite ) , deduced from scattering amplitudes in the bethe - salpeter approach and its various quasipotential reductions ( see @xcite ) , wave functions of the poincare - invariant quantum mechanics ( as an example see wave functions in the instant form of piqm @xcite ) , and also wave functions calculated in the various statements of inverse scattering problems @xcite .
but independently of the method any wave function can be represented as the following laguerre polynomial expansion @xcite : @xmath96 or in the coordinate representation : @xmath97 here @xmath98 are generalized laguerre polynomials , @xmath99 is an euler gamma function , the dimensional parameter @xmath100 can be related to the deuteron matter radius ( see sec .
the wave function representation as a laguerre polynomial expansion ( [ funcu ] ) is very useful for the calculation of the asymptotic behavior of the form factors .
however , one can avoid such representation and obtain the asymptotic expansion directly for the initial wave function .
generally , at high transferred momentum it is necessary to take into account relativistic corrections in the electromagnetic deuteron structure . in our paper relativistic description of the deuteron
is constructed in the framework of instant form of poincare - invariant quantum mechanics ( piqm ) , developed by authors previously @xcite . in this approach
we present electromagnetic deuteron form factors by analogy with nonrelativistic case ( [ gqgnip ] ) .
corresponding formulas in the relativistic impulse approximation were obtained in our paper @xcite : @xmath101 @xmath102 @xmath103 where @xmath104 are the deuteron wave functions in sense of piqm , @xmath105 @xmath106 are relativistic free two - particles charge , quadrupole and magnetic dipole form factors .
formulas for free form factors are given in @xcite .
the deuteron wave functions in sense of piqm are solutions of eigenvalue problem for a mass squared operator for the deuteron ( see , e.g. @xcite ) : @xmath107 an eigenvalue problem for this operator is coincident with the nonrelativistic schrdinger equation within a second order on deuteron binding energy @xmath108 , the value of which is small ( @xmath109 is an averaged nucleon mass ) .
so the deuteron wave functions in sense of piqm differ from nonrelativistic wave functions by conditions of normalization only . in the relativistic case
the wave functions are normalized with relativistic density of states : @xmath110{s}\,k\,u_l(k)\;,\quad s=4(k^{2}+m^{2})\;.\label{relnorm}\ ] ] nonrelativistic formulas eq .
( [ gqgnip ] ) can be obtained from relativistic ones ( [ gqgrip ] ) in the nonrelativistic limit .
this limiting procedure can be performed in the most natural way in the instant form of piqm .
the reason is that in papers @xcite we have constructed the successful formalism of the instant form of piqm . in the case of other forms of piqm ( point and front forms )
the obtaining of nonrelativistic limit is much more difficult . for obtaining the asymptotic form factors behavior at high transferred momentum in the nonrelativistic and relativistic cases it is necessary to estimate asymptotically double integrals ( [ gqgnip ] ) and ( [ gqgrip ] ) at @xmath111 .
notice that integrands reach its maximum value at the integration domain bound , and this point is not a point of extremum . in the previous section the theorem defining asymptotics of @xmath4-tuple integrals of such kind was proven .
we start asymptotic expansion of the deuteron form factors from the nonrelativistic case .
it is caused by the simplicity of the nonrelativistic formulas , so the calculation of the asymptotics is more clear . in
what follows the relativistic calculation will be presented analogous to the nonrelativistic one , although more combersome .
moreover nonrelativistic calculation is interesting because nonrelativistic formulas for the form factors ( [ gqgnip ] ) are conventional , that is why its correct asymptotic expansion has universal significance .
let us emphasize also that the relativistic expressions for form factors and , therefore , their asymptotic estimations depend on the choice of the method of relativisation of the two - nucleon model .
nonrelativistic calculation is also of interest because it helps to clarify the role of relativistic effects in the electromagnetic structure of the deuteron at the asymptotical momentum transfers . as we have seen in sec . 3
, the deuteron form factors in the nonrelativistic impulse approximation can be represented by double integrals ( [ gqgnip ] ) .
we will find its asymptotic expansion using the theorem of section 2 and use as an example the asymptotics of the charge form factor .
we shall estimate only the @xmath112 term in the sum ( [ gqgnip ] ) because the asymptotics of the other terms of form factors ( [ gqgnip ] ) can be derived analogously .
let us rewrite the corresponding @xmath112 term of the charge form factor ( [ gqgnip ] ) using eq .
( [ funcu ] ) : @xmath113 \,k^2\,dk\,k'\,^2\,dk'\;\times\ ] ] @xmath114 we have denoted in ( [ gcnew ] ) : @xmath115 the expression for @xmath116 is commonly accepted ( see , e.g. , @xcite ) : @xmath117 \left(g^p_e(q^2)+g^{n}_e(q^2)\right),\label{tg00}\ ] ] @xmath118 are electric form factors of proton and neuteron respectively , @xmath119 is a step function . in the case under consideration
the space dimension @xmath120 is a large positive parameter .
integration domain is determined by @xmath121-functions in eq .
( [ tg00 ] ) and shown in fig.1 .
the location of the point of maximal value of the function @xmath122 can be obtained by analysis of ( [ s ] ) and ( [ tg00 ] ) : @xmath123 .
let us perform the transition to the new basis as we have descripted before .
we perform the shift of the origin of coordinates to the point of maximal value of the function @xmath122 .
then we rotate the obtained frame for the internal normal to the boundary in the new origin to be coincident with the last basis vector of the new frame .
this procedure is illustrated in fig.1 . in the other words
we perform the transition to the new variables in eq .
( [ gcnew ] ) : @xmath124 at this transformation function @xmath125 gets dependence on large parameter @xmath126 : @xmath127 functions @xmath128 , and the boundary of the integration domain satisfy the conditions of theorem 1@xmath129 , 3@xmath129 , 4@xmath129 . the location of the point , that satisfies the conditions ( [ cond1 ] ) , ( [ cond2 ] ) , can be obtained by a simple analysis of the function ( [ s ] ) : @xmath130 .
let us show that this point satisfies the conditions 2@xmath129 of the theorem .
it is obvious , that in the point of maximal value @xmath131 so the condition ( [ cond1 ] ) is satisfied . let us calculate now the @xmath34 matrix from the condition ( [ cond2 ] ) . in our case
the tangent to the domain of integration boundary vector in the point of maximal value is @xmath132 , i.e. the @xmath34 matrix is a number : @xmath133 we note , that @xmath34 is negative - definite , i.e. the point @xmath134 is really the point of maximal value .
so the point ( 0,0 ) satisfies the condition 2@xmath129 of the theorem .
so integral ( [ gcnew ] ) satisfies the requirements of the theorem proven in the sec .
therefore we can apply the asymptotic formula ( [ ocenka ] ) . calculating by analogy the other terms of the sum ( [ gqgnip ] ) we obtain asymptotic expansions of deuteron form factors in the nonrelativistic impulse approximation : @xmath135 @xmath136
@xmath137 @xmath138 with @xmath139 , variables @xmath140 are denoted in ( [ tt ] ) , @xmath106 , @xmath141 , @xmath142 @xmath143 is defined by equalities : @xmath144 let us perform now the calculation of the relativistic asymptotics of deuteron form factors . to estimate asymptotically integrals ( [ gqgrip ] ) we proceed analogously to nonrelativistic case , i.e. we use relativistic analogs of corresponding nonrelativistic formulas ( [ gcnew])-([b11 ] ) .
now the free relativistic charge form factor in ( [ gqgrip ] ) at @xmath112 is given in ref .
@xcite : @xmath145 @xmath146\;,\label{g00}\ ] ] @xmath147 are electric and magnetic form factors of proton and neutron respectively , @xmath148 @xmath149}^{3/2 } } \frac{1}{\sqrt{1+q^2/4m^2}}\;,\ ] ] @xmath150 @xmath151 and @xmath152 are angles of the wigner spin rotation , @xmath153 + \sqrt{ss'}(\sqrt{s } + \sqrt{s'})}\>,\ ] ] @xmath154 where @xmath155 , @xmath156 , @xmath121 is a step function , @xmath157 , @xmath158 to obtain relativistic asymptotic expansion we also perform transition to the new basis ( shift and rotation ) .
the function @xmath122 and the boundary of the integration domain differ from nonrelativistic ones , so it is necessary to perform a special analysis .
in other words , instead of change of variables ( [ tt ] ) we perform the following replacement : @xmath159
then we obtain asymptotic expansion of the relativistic deuteron form factors by analogy with nonrelativistic case : @xmath160 @xmath161 @xmath162{s\,s'}},\ ] ] @xmath163 functions @xmath164 are specified in eq .
( [ relnorm ] ) , @xmath165 , variables @xmath140 are denoted in ( [ ttrel ] ) , @xmath166 are the binomial coefficients .
asymptotic expansions ( [ ryadner ] ) and ( [ ryadrel ] ) are convergent power series in inverse degrees of the parameter @xmath167 with known coefficients . the asymptotic expansion of this type is obtained in this work for the first time .
one can see from formulas ( [ ryadner ] ) and ( [ ryadrel ] ) , that relativistic corrections change the behavior of form factors at high momentum transfer .
in particular , exponential multiplier index is @xmath126 in the nonrelativistic case , but in the relativistic case it is @xmath167 at @xmath111 .
it seems to be a general feature of our relativistic approach to the description of composite systems , in particular , we have obtained the similar result in consideration of asymptotic behavior of the pion form factor in the composite quark model @xcite .
in this section we represent the obtained asymptotical expansions ( [ ryadner ] ) and ( [ ryadrel ] ) in terms of initial wave functions in the left side of eqs .
( [ funcu ] ) , ( [ relnorm ] ) . for this representation
it is necessary to replace functions @xmath168 by functions @xmath169 in ( [ ryadner ] ) and ( [ ryadrel ] ) using ( [ relnorm ] ) , ( [ utildeu ] ) .
keeping the main term on @xmath170 in asymptotic expansions ( [ ryadner ] ) and ( [ ryadrel ] ) one can obtain the next asymptotic formulas in terms of functions @xmath90 and @xmath104 from ( [ funcu ] ) , ( [ relnorm ] ) : @xmath171 @xmath172{s\,s'}}\right|_{{{t=0}\atop { t'=0}}}\;,\quad i = c , q , m\;. \label{anywfrel}\ ] ] let us note , that similar asymptotic representation can be obtained for any finite number of terms in asymptotic expansions ( [ ryadner ] ) , ( [ ryadrel ] ) . in the modern calculations
the deuteron wave functions are usually represented as a discrete superposition of yukawa - type terms ( see , e.g. , @xcite ) : @xmath173 or in the coordinate representation : @xmath174 @xmath175\ ; , \label{parukoord}\ ] ] @xmath176 coefficients @xmath177 , maximal value of the index @xmath178 and @xmath179 are determined by the best fit of corresponding solution of schrdinger equation . the deuteron wave function analytical form ( [ parukoord ] ) results in the right behavior of the wave functions at large distances : @xmath180 the deuteron wave functions behavior at small distances : @xmath181 is provided by imposing the following conditions on coefficients @xmath182 and @xmath183 : @xmath184 let us substitute the wave functions ( [ paru ] ) to ( [ anywfner ] ) and ( [ anywfrel ] ) , and then obtain the first asymptotic terms of the nonrelativistic deuteron form factors : @xmath185 ^ 2 \left(g_e^p(q^2)+g_e^n(q^2)\right)\ ; , \label{gnrc}\ ] ] @xmath186\left[\sum\limits_{j}d_{j}m_j^4\right ] \left(g_e^p(q^2)+g_e^n(q^2)\right)\ ; , \label{gnrq}\ ] ] @xmath187 ^ 2 \left(g_m^p(q^2)+g_m^n(q^2)\right)\;. \label{gnrm}\ ] ] the dimensional parameter @xmath100 can be found from the expression for the deuteron matter radius in our deuteron model : @xmath188 one can substitute wave functions of the form ( [ funcukoord ] ) into this expression .
so formula ( [ rm ] ) specifies an algebraic equation for @xmath100 .
solution of this equation can be found numerically .
it should be pointed out that main terms of expansion of charge and magnetic form factors in ( [ gnrc ] ) - ( [ gnrm ] ) are determined by @xmath122-state of deuteron only .
the @xmath189-wave function gives the contribution to the main term of the quadrupole from factor .
its faster decrease at @xmath111 in comparison to the other form factors is a consequence of a faster decrease of a @xmath189-wave function at small distances in comparison to @xmath122-wave ( [ atorigin ] ) . from the mathematical point of view
the type of leading terms in ( [ gnrc])-([gnrm ] ) is a consequence of conditions on the coefficients ( [ cond ] ) .
the modification of these conditions obviously results in change of the main terms in ( [ gnrc])-([gnrm ] ) . from these formulas
it is also noticed that asymptotic expansions for the deuteron form factors contain dependence on the asymptotics of nucleon form factors .
we emphasize , that in the other deuteron asymptotics investigations only the power dependence on the transferred momentum was calculated as a rule . in the present paper
we give a rigorous calculation of a multiplicative preasymptotical constant .
one can calculate relativistic asymptotics of form factors by analogy with nonrelativistic case . for this calculation
we use the formulas ( [ relnorm ] ) , ( [ anywfrel ] ) , ( [ cond ] ) . as a result
we obtain : @xmath190 @xmath191 notice that asymptotic expansions ( [ gnrc])-([gnrm ] ) and ( [ gcm]),([gq ] ) are obtained for the first time in our work .
it is interesting to compare obtained asymptotic estimations ( [ gnrc ] ) - ( [ gnrm ] ) , ( [ gcm]),([gq ] ) with observable behavior of the deuteron characteristics . at present time there
exists the experimental information about function @xmath192 entered the differential cross section of the elastic @xmath193-scattering .
this function is expressed in terms of the deuteron form factors @xcite .
the values of function @xmath192 are known up to @xmath0 6 ( gev / c)@xmath1 . for the comparison with experimental data one needs to specify asymptotics of the nucleon form factors .
it is naturally to choose for nucleon form factors the asymptotic which is predicted by the quark model @xcite @xmath194 . under these conditions the power dependence on @xmath126 of the function @xmath192 coincides with
experimentally observed one .
the physical consequences will be examined in detail in the other paper .
the theorem defining asymptotics of multiple integrals of some special type is proved . with help of the proven theorem
the asymptotic expansion of the deuteron electromagnetic form factors at @xmath3 is calculated for the first time .
the expansion is represented as a convergent series on in inverse powers of momentum transfer .
the asymptotic of the form factors is found in terms of the conventional representation of the deuteron wave function as a discrete superposition of yukawa - type terms .
the asymptotic behavior of the form factors is calculated in the nonrelativistic impulse approximation and in the relativistic invariant impulse approximation proposed by the authors in the instant form of the poincare - invariant quantum mechanics previously .
it is established that relativistic corrections change the power dependence of the form factors on the momentum transfer at @xmath3 and slow down its decrease .
it is also found that relativistic effects result in the agreement of the theoretical asymptotics and the experimentally observed behavior of the structure function @xmath192 at highest achieved momentum transfers .
this work was supported in part by russian foundation of basic researches ( grant 07 - 02 - 00962 ) . | a theorem about asymptotic estimation of multiple integral of a special type is proved for the case when the integrand peaks at the integration domain bound , but not at a point of extremum . using this theorem the asymptotic expansion of the electromagnetic deuteron form factors at high momentum transfers
is obtained in the framework of two - nucleon model in both nonrelativistic and relativistic impulse approximations .
it is found that relativistic effects slow down the decrease of deuteron form factors and result in agreement between the relativistic asymptotics and experimental data at high momentum transfers . * asymptotic estimation of some multiple integrals and the electromagnetic deuteron form factors at high momentum transfer * a.f .
krutov _ department of general and theoretical physics , _ _ samara state university , _ _ samara , _ _ 443011 russia _ v.e .
troitsky _ d.v .
skobeltsyn institute of nuclear physics , _ _ moscow state university , _ _ moscow , _ _ 119992 russia _ n.a
. tsirova _ department of general and theoretical physics , _
_ samara state university , _ _ samara , _ _ 443011 russia _ ( dated : 10 september 2007 ) pacs numbers : 11.10.jj , 13.40gp , 13.75.cs |
a hyperelliptic function field over @xmath3 is a function field @xmath2 with genus @xmath80 that contains a rational subfield @xmath81 with @xmath82=2 $ ] 1 . a function field @xmath2 of genus @xmath80 is hyperelliptic if and only if there is a divisor @xmath83 with @xmath84 and the dimension of the riemann space at @xmath85 is greater or equal than 2 2 .
every @xmath2 of genus 2 is hyperelliptic every element @xmath74\in{\ensuremath{\mathbb{j}_{\mathbb{f}}}}$ ] of genus 2 with @xmath90 , this divisor can be represented by the pair of functions @xmath91 such that @xmath92 , @xmath93 and @xmath94 with @xmath95 monic , @xmath96 and @xmath97 .
the pair @xmath91 is called * mumford representation * of @xmath74 $ ] . * artin s aproximation theorem * @xcite + let @xmath2 a function field and @xmath98 different places in pairs of @xmath2 , @xmath99 and @xmath100 then there exists @xmath101 such that : + + @xmath102 with @xmath103 this theorem generalize the chinese remainder theorem and it will assure the existence of a curve that passes through the given points ( places ) with any degree of multiplicity at the hyperelliptic curve ( valuation at the point ) we just justify the structure of a hyperelliptic function field of genus two , we have the function field @xmath104 is such that @xmath105 with @xmath106 .
+ so we define the hyperelliptic curve as @xmath107 + + given two divisors @xmath108 and @xmath109 , we want to find the divisor class @xmath110\oplus [ { \ensuremath{\mathcal{p}}}_{1}'+{\ensuremath{\mathcal{p}}}_{2}'-2{\ensuremath{\mathcal{q}}}_{\infty}]$ ] , to find this , we can use the approximation theorem to be sure of its existence , we have that there is a function @xmath111 and its principal divisor @xmath112 that has in the support the places of degree one @xmath113 , for this , we have to find an interpolation polynomial that passes through this places , so we make this polynomial equal to @xmath105 @xmath114 and we solve for @xmath115 and finally we make hyperelliptic involution to find @xmath116=[{\ensuremath{\mathcal{p}}}_{1}''+{\ensuremath{\mathcal{p}}}_{2}''-2{\ensuremath{\mathcal{q}}}_{\infty}]$ ] .
+ the geometric intuition of what we want to find ( blue ) given two divisors with disjoint supports : + using mumford representation , if @xmath75 is a reduced divisor as in the corresponding theorem , and we denote the places as @xmath117 , then @xmath118 with @xmath119 and @xmath120 for all @xmath117 so @xmath121 and @xmath122 , so the addition is done as following : + + @xmath123 + @xmath124 + + we want to find @xmath125 as in the previous figure to represent @xmath126 , this results inverting @xmath127 and @xmath128 + + to find the other elements of @xmath129 @xmath127 and @xmath130 we have to find the interpolation polynomial for the given places , square it and then making it equal to @xmath105 : + + we have that : + @xmath131 + + if we solve : + + @xmath132 mod @xmath133 + @xmath134 mod @xmath135 + + with @xmath136 we will have : + + @xmath137 + @xmath138 + + so @xmath139 and we will have a @xmath140 system of equations , the solutions are going to be the coefficients @xmath141 of @xmath142 , if we do the calculations reducing @xmath143 modulo @xmath133 and @xmath143 modulo @xmath135 we can find that @xmath144 in general induces a matrix + + @xmath145 mod @xmath146 + + @xmath147 mod @xmath148 + + this induces 4 equations : + + @xmath149 + @xmath150 + @xmath151 + @xmath152 + + as we know the values @xmath153 and we want @xmath154 to find the coefficients of @xmath142 the system is this : + @xmath155.\ ] ] the solution of this matrix give us the coefficients @xmath141 of @xmath142 having this we just have to make it equal to the hyperelliptic curve @xmath105 : + @xmath156 + this happens because @xmath157 and it has @xmath95 and @xmath158 as factors and the coordinates of @xmath61 are roots of the places of @xmath159 and @xmath160 so @xmath161 is the polynomial of degree 2 resulting of the quotient and this is going to have as roots the coordinates of @xmath61 in @xmath162 .
+ to find @xmath163 we have that @xmath164 we just have to evaluate the roots of @xmath161 in the hyperelliptic curve @xmath165 over @xmath166 but this can be a problem ( computing roots ) so another way is to check that : + + @xmath167 + with this we have @xmath168\oplus [ d_{2}]= [ d_{3}]=(u''(x),v''(x))$ ] .
see @xcite but here we will show a slightly different approach , which in this case is a particular case of the previous , suppose we have @xmath169 and we want to calculate @xmath170 $ ] , this divisor is linearly equivalent to a divisor with * prime places of degree two * : @xmath170 \sim [ 2\mu -2\infty ] \oplus [ 2\omega - 2\infty]$ ] what we have here is @xmath171 and @xmath172 , the geometric intuition is that we have two points which are tangent to the hyperelliptic curve with degree of intersection two in both : + the mumford divisors are the following : + + let @xmath173 and @xmath174 be two points ( places ) of the hyperelliptic curve @xmath0 of genus two , @xmath169 , we will calculate + + @xmath170 = [ d_1]\oplus [ d_2 ] = [ 2[\mu-2\infty]\oplus [ 2\omega-2\infty]$ ] + + @xmath175=<x^2 - 2x\mu_x+{\mu_x}^2,\frac{dc}{dx}(\mu_x,\mu_y)x-\frac{dc}{dx}(\mu_x,\mu_y)\mu_x+\mu_y>$ ] + @xmath176=<x^2 - 2x\omega_x+{\omega_x}^2,\frac{dc}{dx}(\omega_x,\omega_y)x-\frac{dc}{dx}(\omega_x,\omega_y)\omega_x+\omega_y>$ ] + + here we built the quadratic polynomial as a double root in the @xmath61 coordinate of the prime place of each divisor , and we used formal differentiation for the linear part to get the tangent line to the curve @xmath0 at the given point ( place ) .
+ + so , using the matrix in the case 1 we need to solve for @xmath177 such that @xmath178 : + @xmath179 + @xmath180 + @xmath181 + @xmath182 + + @xmath183 + @xmath184 + @xmath185 + @xmath186 @xmath187.\ ] ] in the same way with the solution of this system with get the coefficients of @xmath142 and the new places to do hyperelliptic involution , so we are ready to compute @xmath170=(u'(x),v'(x))$ ] + in this new case we have that both divisors share a place @xmath188 , let @xmath189 and @xmath190 , we will need to detect explicitly the repeated place @xmath36 to calculate this case given the divisors in the mumford notation , but this is easy because if @xmath168=<u_1(x)=x^2+ax+b , v_1(x)>$ ] and @xmath191=<u_2(x)=x^2+\alpha x+\beta , v_2(x)>$ ] the matrix generated by case 1 must be singular , so @xmath192 and @xmath193 have a common root so @xmath194 implies that the repeated root @xmath61 is @xmath195 , having this we have that @xmath196 and the other places @xmath197 and @xmath198 can be calculated directly , so we are ready to calculate the addition in this case .
we see here the place @xmath36 of degree 2 ( blue curve tangent point at @xmath0 ) .
+ + so we rewrite @xmath199 and @xmath200 as : + @xmath201 + @xmath202 + + and we have the mumford divisors in the following way : + + @xmath175=<x^2 - 2sx+s^2,\frac{dc}{dx}(s , t)x-\frac{dc}{dx}(s , t)s+t>$ ] + @xmath176=<x^2-x(\mu_x+\omega_x)+\mu_x\omega_x,\frac{\mu_{y}-\omega_{y}}{\mu_{x}-\omega_{x}}x+\frac{\mu_{x}\omega_{y}-\omega_{x}\mu_{y}}{\mu_{x}-\omega_{y}}>$ ] + + here we have that @xmath175 $ ] has a multiple root at @xmath203 and the linear part is the tangent line at @xmath204 over @xmath0 and @xmath176 $ ] the linear part is the line through @xmath205 and @xmath206 , and the quadratic part has @xmath207 and @xmath208 as roots .
+ we use the case 1 and we have that : + as in the other cases we use the matrix of case 1 to obtain the coefficients of the interpolation polynomial @xmath111 with @xmath105 , @xmath106 and @xmath178 .
so we solve for @xmath177 the following matrix : + @xmath209.\ ] ] jacobians of hyperelliptic curves of genus 2 are a good candidate for asymmetric cryptography , so optimization of its endomorphism is an important task , the work of costello @xcite is very important because it shows that the calculation of the addition in @xmath210 can be done solving a system of linear equations over the base field in two cases given mumford divisors of a genus 2 curve jacobian , we have extended with a similar approach a third case when both divisors to add in the jacobian share a point , the existence of the solution is backed up by the artin s approximation theorem , the figures shown in this work were calculated using these formul . siegfried bosch _ algebraic geometry and commutative algebra_. + springer , utx 2013 .
niederreiter and xing , _ rational points on curves over finite fields_. cambridge university press , lms 285 , 2001 .
blake , seroussi , smart , _ elliptic curves in cryptography_. + cambridge university press , lms 265 , 1999 .
thomas wollinger , _ software and hardware implementation of hyperelliptic curve cryptosystems_. + ruhr - universit"at bochum , it security 1 , 2004 .
tanja lange _ formulfor arithmetic on genus 2 hyperelliptic curves_. + springer - verlag applicable algebra in engineering , communication and computing , vol 15 , num 5 2005 .
henning stichtenoth _ algebraic function fields and codes_. + springer , graduate texts in mathematics , 2009 .
neal koblitz _
algebraic aspects of cryptography_. + springer , algorithms and computation in mathematics , 1999 .
klaus hulek _ elementary algebraic geometry_. + ams , student mathematical library 2003 .
whitfield diffie , martin hellman _
new directions in cryptography_. + ieee , ieee transactions on information theory 22 1976 .
david g. cantor _ computing in the jacobian of a hyperelliptic curve_. + ams , mathematics of computation 48 , ams 1987 .
craig costello , kristin lauter _ group law computations on jacobians of hyperelliptic curves_. + springer - verlag , iacr sac11 proceedings of the 18th international conference on selected areas in cryptography 2011 .
julio lpez and ricardo dahab _ an overview of elliptic curve cryptography _ + pelzl , jan _ hyperelliptic cryptosystems on embedded microprocessors _ + communication security group ,
rhr - universitt bochum | we present an efficient endomorphism for the jacobian of a curve @xmath0 of genus 2 for divisors having a non disjoint support .
this extends the work of costello in @xcite who calculated explicit formulfor divisor doubling and addition of divisors with disjoint support in @xmath1 using only base field operations .
explicit formulis presented for this third case and a different approach for divisor doubling .
high speed implementation of asymmetric cryptosystems is an important requirement to secure communications using devices with a reduced processor power , commonly rsa is used to provide asymmetric cryptography but this is not always the best solution for public key cryptography because of the high processing requirements for big random prime generation .
+ elliptic curve cryptography ( ecc ) provides shorter keys and faster computation than rsa in some platforms and they offer the same security with less bits in the key @xcite .
the jacobian of a hyperelliptic curve is an excellent candidate for discrete logarithm problem based cryptography with the same safety benefits as rsa but with even shorter bit - lengths keys @xcite .
hyperelliptic curve cryptography ( hecc ) has been studied by lange , wollinger @xcite , @xcite to get explicit formulfor the calculation of the group operation over the jacobian of genus 2 hyperelliptic curves , there are algorithms for arbitrary genus @xcite , we will present a method to calculate the group operation using only arithmetic in the base field without polynomial pseudo inversion ( lange ) using a geometrical approach for genus 2 hyperelliptic function fields using divisor theory and the mumford representation of the elements of the jacobian of the hyperelliptic function field , this approach has been studied by @xcite solving a system of linear equations over the base field to get the coefficients of the polynomial that defines the addition of two divisors in two cases : point doubling and regular addition , but regular addition has a subcase when the pair of divisors to be added share a point ( place ) in their supports , this third case is presented using formal differentiation with explicit formul .
following @xcite , an algebraic function field @xmath2 in one variable over @xmath3 is an extension @xmath4 such that @xmath5 is a finite algebraic extension of @xmath6 for some transcendental @xmath7 over @xmath3 .
a * valuation ring * in the function field @xmath2 is a ring @xmath8 such that @xmath9 and for all @xmath10 , @xmath11 or @xmath12 .
these rings are * local rings * i.e. they have only one maximal ideal @xmath13 , where @xmath14 are the units of @xmath15 and all the ideals of @xmath15 are principal .
it follows that every @xmath16 has a unique representation @xmath17 for some @xmath18 and @xmath19 .
a * place * @xmath20 of @xmath2 is the maximal ideal of a valuation ring @xmath15 of @xmath2 .
if @xmath15 is a valuation ring of @xmath2 and @xmath20 is its maximal ideal , then @xmath15 is determined only by @xmath20 and we denote this ring by : + + @xmath21 + + here we say that @xmath22 is the valuation ring at the place @xmath20 and the number of elements of @xmath23 is infinite @xcite .
let @xmath24 , the map @xmath25 is a discrete valuation of @xmath2 associated to @xmath20 in the way that if @xmath26 is a * uniformization variable * then for all @xmath27 exists a unique representation of @xmath28 , @xmath17 with @xmath29 and @xmath18 , then we define @xmath30 and @xmath31 @xmath32 will be the field of residual classes of @xmath20 , ( this is field because @xmath20 is maximal in @xmath33 , this classes will be defined as @xmath34 . @xmath35 $ ] will be the degree of @xmath36 .
it follows that @xmath37 is finite .
let @xmath38 and @xmath24 , we say that @xmath20 is a zero of @xmath28 if @xmath39 and @xmath20 is a pole of @xmath28 if @xmath40 this is that if @xmath41 then @xmath20 is a zero of @xmath28 of order m , if @xmath42 we say that @xmath20 is a pole of @xmath28 of order @xmath43 .
every @xmath38 transcendental over @xmath3 has at least one zero and one pole , in fact it has the same finite number of zeroes and poles .
a divisor is a finite formal sum of places @xmath44 the support of a divisor @xmath45 is defined as @xmath46 given @xmath47 y @xmath48 the sum is done coefficient wise : @xmath49 the zero element of @xmath50 is : @xmath51 for @xmath52 and @xmath45 we define @xmath53 , then @xmath54 a partial order in the group of divisors is given by @xmath55 a divisor such that @xmath56 is * positive * or * effective * .
the * degree of a divisor * is defined as @xmath57 for a function @xmath58 let @xmath59 be the set of zeros and @xmath60 be the set of poles of @xmath61 in @xmath62 .
we define @xmath63 @xmath64 @xmath65 divisors @xmath66 and @xmath67 are effective divisors , and @xmath68 the set @xmath69 is called the * subgroup of principal divisors * of @xmath70
. the quotient group @xmath71 will be defined as * the group of divisor classes or jacobian of @xmath2*. for @xmath72 , the corresponding element in @xmath73 is denoted by @xmath74 $ ] , the class of @xmath75 .
two divisors @xmath75 , @xmath76 are equivalent ( @xmath77 ) if @xmath74=[d']$ ] , this is that , @xmath78 for some @xmath79 .
this is an equivalence relation . |
in this paper , we study mapping classes defined by bipartite coxeter graphs with sign - labels on the vertices determined by the bipartite structure .
if the graph is connected and has at least two vertices , then these _ alternating - sign coxeter mapping classes _ are pseudo - anosov , and if the coxeter graph is a tree the associated mapping class is the monodromy of an alternating fibered knot or link , which we call an _ ( alternating ) coxeter link_. there has long been interest in the location of roots of alexander polynomials for alternating links .
murasugi showed that the coefficients of the polynomials have alternating signs , and hence no real root can be negative @xcite .
hoste conjectured that the real part of all zeros must be bounded from below by @xmath0 .
this and related conjectures were settled for some classes of alternating links in @xcite . using properties of alternating - sign coxeter transformations , we give a simple proof that the roots of the alexander polynomials for alternating coxeter links are real and positive . by a result of perron and rolfsen @xcite
, this implies that the fundamental group of the complement of an alternating coxeter link is bi - orderable .
applying an interlacing property for alternating - sign coxeter graphs , we show that the homological dilatations are monotone under graph inclusion .
thus the minimum homological dilatation achieved by an alternating coxeter link is @xmath1 , the square of the golden ratio .
similar properties hold for the alexander polynomial of the mapping torus of alternating - sign coxeter mapping classes . _ in @xcite hirasawa and murasugi similarly study the roots of alexander polynomials for quasi - rational knots and links , which include the coxeter links discussed in this paper , and they also prove stability and interlacing properties of the alexander polynomial for these examples . by applying the constructs of coxeter graphs and coxeter transformations in this paper , we simplify their proofs in this context , and extend the results to more general mapping classes and mapping tori associated to alternating - sign coxeter graphs .
_ the alexander polynomial @xmath2 $ ] is an invariant of a finitely presented group with a prescribed homomorphism onto @xmath3 .
given a knot or link @xmath4 in @xmath5 , each oriented seifert surface @xmath6 defines a surjective homomorphism of @xmath7 to @xmath3 by algebraic intersection of closed paths with @xmath6 .
denote by @xmath8 the associated alexander polynomial .
if @xmath9 is fibered over the circle with fiber @xmath6 and monodromy @xmath10 , then @xmath8 is the characteristic polynomial of the homological monodromy @xmath11 ( this can be deduced from either the fox calculus or the seifert algorithm for finding @xmath12 , see e.g. @xcite ) . given any mapping class @xmath10 on a surface @xmath6 , write @xmath13 for the characteristic polynomial of the homological monodromy .
it follows that if @xmath4 is a fibered link with monodromy @xmath14 , and @xmath15 is the alexander polynomial of @xmath4 , we have @xmath16 there are few restrictions on the alexander polynomial : any monic , reciprocal polynomial can be realized as @xmath13 up to multiples of @xmath17 and @xmath18 , where @xmath14 is the monodromy of some fibered link @xcite .
the story is different when we confine ourselves to _ alternating knots and links _ : those that admit a planar projection such that over and under crossings are alternating .
murasugi showed in @xcite that if @xmath6 is the seifert surface defined by an alternating planar projection , then @xmath19 has degree @xmath20 , and the coefficients for the powers @xmath21 are all strictly positive or strictly negative for @xmath22 .
this implies , for example , that any real root of @xmath23 must be positive . in 2002 , hoste conjectured the following : for alternating knots , the real part of any zero of the alexander polynomial is strictly greater than @xmath0
. a lower bound on the real part of roots of @xmath23 was found by lyubich and murasugi @xcite for two - bridge links .
the results were later improved by koseleff and pecker @xcite , and stoimenow @xcite .
hirasawa and murasugi in @xcite showed that for a large class of alternating links , the roots of the alexander polynomial are real and positive , a property of integer polynomials known as _ real stability_. our first result is the following .
[ realroots - thm ] if @xmath14 is an alternating - sign coxeter mapping class , then @xmath24 has real stability . in particular , the alexander polynomial of an alternating coxeter link has real stability .
fox s _ trapezoidal conjecture _ concerns the coefficients of alexander polynomials of alternating knots .
let @xmath25 be the alexander polynomial of an alternating knot .
then there exists an integer @xmath26 satisfying @xmath27 such that @xmath28 the trapezoidal conjecture has been verified for several classes of alternating knots , e.g. for algebraic alternating knots by murasugi @xcite and alternating knots of genus two by ozsvth and szab @xcite and jong @xcite .
real stability implies the trapezoidal property for integer polynomials .
the coefficient sequence of a polynomial @xmath29 $ ] with only positive real roots is strictly _ log - concave _ , i.e. @xmath30 holds for all @xmath31 , see e.g. @xcite .
thus , the trapezoidal property of alexander polynomials of alternating coxeter links follows from theorem [ realroots - thm ] ( cf .
more generally , we have the following .
[ trapezoidal - thm ] if @xmath14 is an alternating - sign coxeter mapping class , then @xmath24 is trapezoidal . in particular , alternating - sign coxeter links have trapezoidal alexander polynomials .
a second application of theorem [ realroots - thm ] is the bi - orderability of knot groups and fundamental groups of 3-manifolds .
a group @xmath32 is _ bi - orderable _ if it admits a total order @xmath33 on @xmath32 that is compatible with the group operation , that is @xmath34 perron and rolfsen showed that if all the eigenvalues of the homological action of a surface homeomorphism @xmath10 are real and positive , then the fundamental group of its mapping torus is bi - orderable @xcite .
thus , theorem [ realroots - thm ] has this immediate consequence .
[ biorderable - thm ] the mapping torus of an alternating - sign coxeter mapping class has bi - orderable fundamental group .
a _ mapping class _ on an oriented compact surface @xmath6 of finite type is a self - homeomorphism up to isotopy relative to the boundary .
the _ homological dilatation _
@xmath35 of a mapping class @xmath10 is the largest eigenvalue ( in modulus ) of the characteristic polynomial of the action of @xmath10 on first homology . by the nielsen - thurston classification theorem
, mapping classes fall into three types : those that are periodic , non - periodic but preserving the isotopy class of a simple closed multi - curve , and _ pseudo - anosov_. the third type is the most general , and has the property that for some pair of transverse measured singular foliations @xmath36 , the mapping class stretches the measure @xmath37 by @xmath38 and @xmath39 by @xmath40 for some @xmath41 .
the constant @xmath42 is the _ ( geometric ) dilatation _ of the mapping class .
the homological and geometric dilatations are related as follows @xmath43 with equality if and only if @xmath10 is _ orientable _ , i.e. , its invariant foliations @xmath44 are orientable ( see , e.g. @xcite ) .
the mapping torus of a mapping class @xmath14 is the 3-dimensional manifold @xmath45/(x,1 ) \sim ( \phi(x),0).\ ] ] by a theorem of w. thurston , this manifold admits a hyperbolic structure if and only if @xmath10 is pseudo - anosov @xcite .
the associated fibration @xmath46 defines a surjective homomorphism @xmath47 and a corresponding alexander polynomial @xmath24 .
we show that the dilatation of alternating - sign coxeter mapping classes is monotonic with respect to graph inclusion .
thus the minimum dilatation for alternating - sign coxeter mapping classes is achieved by the alternating - sign @xmath48 graph , which in turn is geometrically realized by the figure eight knot .
[ goldenmean - thm ] the minimum homological and geometric dilatation of alternating - sign coxeter mapping classes is the square of the golden ratio @xmath1 , and is geometrically realized as the monodromy of the figure eight knot .
[ lehmer - rem ] _ by a result of mcmullen @xcite the spectral radius of the classical coxeter transformations is minimized by the @xmath49 coxeter graph , also known as the @xmath50 star - like graph @xcite .
the associated coxeter link is the @xmath51-pretzel link @xcite and the dilatation of its monodromy is the conjectural smallest salem number , known as lehmer s number @xcite , which is smaller than the square of the golden ratio . _
_ by contrast to theorem [ goldenmean - thm ] , when dropping the assumption of alternating signs , it is possible to find mixed - sign coxeter graphs whose associated mapping classes have dilatation arbitrarily close to 1 ( see @xcite ) . _ in section [ coxeter - sec ] we recall some definitions and properties of classical coxeter systems and generalize them to mixed - sign coxeter systems .
the analog of alexander polynomials for coxeter systems is the coxeter polynomial , the characteristic polynomial of the coxeter transformation .
for bipartite alternating - sign coxeter systems , we prove real stability for the coxeter polynomial and the interlacing property .
section [ geometry - sec ] discusses geometric realizations of alternating - sign coxeter systems and contains proofs of theorems [ realroots - thm ] and [ goldenmean - thm ] .
* acknowledgements .
* the authors are grateful to n. acampo , s. baader , k. murasugi and to the anonymous referee for their helpful comments and suggestions .
a _ mixed - sign coxeter graph _ is a pair @xmath52 , where @xmath53 is a finite connected graph without self- or double edges and @xmath54 is an assignment of a sign @xmath55 or @xmath56 to every vertex @xmath57 of @xmath53 .
let @xmath58 be the vector space of @xmath59-labelings of the vertices of @xmath53 .
for @xmath60 , let @xmath61 $ ] be the corresponding element of @xmath58 giving the label @xmath62 on @xmath63 and @xmath64 on all other vertices of @xmath53 .
the real vector space @xmath65 is equipped with a symmetric bilinear form @xmath66 , given by @xmath67 , [ v_i ] ) = -2\cdot\mathfrak{s}(v_i)$ ] and otherwise @xmath67 , [ v_j ] ) = a_{ij}$ ] , where @xmath68 is the adjacency matrix of @xmath53 . to every vertex @xmath57 , we associate a reflexion @xmath69 about the hyperplane of @xmath65 perpendicular to @xmath70 $ ] , given by the formula @xmath71 ) = [ v_j ] - 2\frac{b([v_i],[v_j])}{b([v_i],[v_i])}[v_i].\ ] ] the _ coxeter transformation _ is the product @xmath72 of all these reflections . for trees ,
this product does not depend , up to conjugation , on the order of multiplication @xcite , but in general it does . for bipartite coxeter graphs
@xmath53 , however , there is a distinguished conjugacy class , the _ bipartite coxeter transformation _ @xmath73 given by @xmath74 , where @xmath75 is any product of all the reflections corresponding to vertices in one part of the partition and @xmath76 is any product of all the reflections corresponding to vertices in the other part .
this is well - defined since all the reflections corresponding to vertices in one part of the partition commute pairwise . if all signs @xmath54 of a bipartite coxeter graph are positive , theorems of acampo and mcmullen state that the eigenvalues of the bipartite coxeter transformation are on the unit circle or positive real and that the spectral radius is monotonic with respect to graph inclusion @xcite .
we now prove analogs of these theorems for _ alternating - sign coxeter graphs _
, the case where the bipartition of the graph @xmath53 is actually given by the signs @xmath54 .
[ bicolour - prop ] let @xmath52 be an alternating - sign coxeter graph .
then the eigenvalues of the bipartite coxeter transformation @xmath73 are real and strictly negative .
let @xmath52 be an alternating - sign coxeter graph .
number the vertices of @xmath53 starting with all the positive ones , and then proceeding to the negative ones . with this vertex numbering ,
the adjacency matrix @xmath77 of @xmath53 becomes a @xmath78-block matrix with zero blocks on the diagonal and blocks @xmath79 and @xmath80 in the upper right and lower left , respectively . using the above formula for the @xmath69
, we have that the products @xmath75 and @xmath76 corresponding to the partition are given by @xmath81 multiplication of @xmath75 and @xmath76 shows that the bipartite coxeter transformation @xmath82 is symmetric .
therefore , @xmath73 has only real eigenvalues .
it is left to show that there are no positive eigenvalues .
note that @xmath83 .
furthermore , by expanding we obtain @xmath84 and thus , for any eigenvalue @xmath85 of @xmath73 , we have @xmath86 where @xmath87 is some eigenvalue of the adjacency matrix @xmath88 .
it follows that @xmath89 is a non - positive real number , since @xmath87 is a real number .
in particular , every eigenvalue @xmath38 of the alternating - sign coxeter transformation @xmath73 is strictly negative .
let @xmath53 and @xmath90 be alternating - sign coxeter graphs so that @xmath53 is a subgraph of @xmath90 .
we say that @xmath90 is obtained from @xmath53 by a _
vertex extension _ if the vertex set of @xmath90 contains one more element @xmath91 than the vertex set of @xmath53 , and the edges of @xmath53 are precisely the edges of @xmath90 that do not have @xmath91 as an endpoint .
[ interlacing - prop ] let @xmath52 and @xmath92 be two alternating - sign coxeter graphs . if @xmath90 is a vertex extension of @xmath53 , then the eigenvalues of the bipartite coxeter transformations @xmath73 and @xmath93 are interlaced , i.e. , if @xmath94 are the eigenvalues of @xmath95 , and @xmath96 are the eigenvalues of @xmath97 , then @xmath98 let @xmath52 be an alternating - sign coxeter graph with bipartite coxeter transformation @xmath73 . from the proof of proposition [ bicolour - prop ] ,
we recall that the eigenvalues of @xmath73 are in one - to - one correspondence with the eigenvalues of the adjacency matrix @xmath88 .
more precisely , the correspondence is given by @xmath99 where @xmath38 and @xmath87 are eigenvalues of @xmath73 and @xmath88 , respectively . since @xmath53 is bipartite , the eigenvalues of @xmath88 are symmetric with respect to the origin @xcite .
furthermore , since @xmath100 is monotonically increasing with respect to @xmath101 , there exists a monotonic transformation of @xmath59 taking the eigenvalues of @xmath88 to the eigenvalues of @xmath73 .
now let @xmath92 be an alternating - sign coxeter graph with bipartite coxeter transformation @xmath93 such that @xmath90 is a vertex - extension of @xmath53 .
then the eigenvalues of @xmath88 and @xmath102 are interlaced @xcite and therefore so are the eigenvalues of @xmath73 and @xmath93 .
[ minimum - prop ] the minimum spectral radius for an alternating - sign coxeter transformation is realized by the alternating - sign @xmath48 coxeter graph , and the spectral radius is the square of the golden mean . noting that every non - trivial alternating - sign coxeter graph is a ( perhaps multiple ) vertex extension of the alternating - sign @xmath48 graph , the statement follows from proposition [ interlacing - prop ] .
_ if a bipartite graph @xmath53 is a subgraph of another bipartite graph @xmath90 with one more vertex but @xmath90 is not a vertex - extension of @xmath53 , then the eigenvalues of the corresponding adjacency matrices need not be interlaced . _ choosing @xmath53 and @xmath90 as in figure [ non - interlacing ] ,
the eigenvalues of the adjacency matrix of @xmath53 are given by @xmath103 and the eigenvalues of the adjacency matrix of @xmath90 are given by @xmath104 .
in particular , these eigenvalues are not interlaced . however , focusing on the largest eigenvalue , it is still true that the spectral radius is monotonic under graph inclusion .
[ monoton - prop ] let @xmath52 and @xmath92 be two alternating - sign coxeter graphs .
if @xmath53 is a subgraph of @xmath90 , then the spectral radius of @xmath73 is less than or equal to the spectral radius of @xmath93 .
the proof is basically the same as the proof of proposition [ interlacing - prop ] .
however , instead of interlacing ( which does not necessarily apply in the case of non - induced subgraphs ) , we use perron - frobenius theory and the fact that @xmath88 is dominated by a submatrix of @xmath102 .
_ general coxeter graphs are defined with arbitrary edge weights @xmath105 .
the corresponding entries @xmath106 of the adjacency matrix are then defined to be @xmath107 . although we formulated propositions [ bicolour - prop ] and [ monoton - prop ] for constant edge - weights @xmath108 ,
they also hold in this generalized context .
proposition [ bicolour - prop ] holds without change of wording . for proposition [ monoton - prop ] , we must add the assumption that when @xmath53 is a subgraph of @xmath90 , then every edge - weight of @xmath53 is less than or equal to the edge - weight of @xmath90 . _
in this section , we associate fibered alternating links and more general mapping tori to alternating - sign coxeter graphs @xmath109 . as in the _ classical _ ( or _ positive - sign _ ) case , a mixed - sign coxeter graph with @xmath110 vertices defines a subgroup of the general linear group @xmath111 generated by reflections . in the classical case ,
the reflections preserve an associated symmetric bilinear form @xmath112 , where @xmath113 is the adjacency matrix of the coxeter graph .
for a mixed - sign coxeter system the bilinear form is given by @xmath114 , where @xmath115 is a diagonal matrix with @xmath116 entries on the diagonal depending on the signs @xmath117 assigned to vertices of the coxeter graph .
for mixed - sign coxeter graphs , just as for classical ones , one can explicitly construct mapping classes whose homological monodromy is conjugate to the coxeter transformation up to sign @xcite .
classical bipartite coxeter systems have been shown to have many useful properties .
acampo showed that all eigenvalues of the coxeter transformation are real or lie on the unit circle .
this condition is sometimes called _ bi - stability _ @xcite .
since the traces of the eigenvalues over the reals are related to the eigenvalues of the adjacency matrix of the coxeter graph , the eigenvalues satisfy an interlacing theorem .
mcmullen used this to prove monotonicity of the spectral radius of coxeter transformations with respect to graph inclusion , and found a sharp lower bound for the gap between 1 and the next smallest spectral radius of coxeter transformations @xcite .
it follows , in particular , that the classical coxeter mapping classes associated to bipartite classical coxeter graphs that are not spherical or affine have dilatation bounded from below by lehmer s number , which is approximately 1.17628 @xcite .
[ positivity - rem ] _ by contrast to theorem 3 , acampo showed that for any classical bipartite coxeter graph that is not spherical or affine , the roots of the corresponding coxeter polynomials are either on the unit circle or positive real , with at least one root greater than 1 @xcite .
if hoste s conjecture is true , this gives a homological proof of the fact that the knots associated to classical bipartite coxeter graphs that are not spherical or affine can never be alternating .
this can also be proved independently : such a knot is positive , i.e. it has a diagram with only positive crossings . for the signature @xmath118 and genus @xmath119
, we have @xmath120 , since @xmath20 equals the number of vertices and @xmath118 equals the signature of the bilinear form @xmath112 .
but for knots which are both positive and alternating , @xmath121 holds , e.g. by properties of rasmussen s @xmath122-invariant @xcite .
_ let @xmath123 be an arrangement of line segments in the plane whose intersection graph equals @xmath53 .
that is , to each vertex @xmath63 of @xmath53 there is an associated line segment @xmath124 in @xmath123 , and two line segments in @xmath123 intersect if the corresponding vertices are connected by an edge of @xmath53 .
a _ planar realization of _ @xmath53 is an embedding of @xmath125 in @xmath126 with coordinate axes @xmath127 and @xmath128 , so that if @xmath129 , then @xmath124 is parallel to the @xmath128-axis , and if @xmath130 , then @xmath124 is parallel to the @xmath127-axis .
if @xmath53 has a planar realization , then we thicken the @xmath124 into rectangular strips @xmath131 $ ] ( resp . , @xmath132 \times \ell_v$ ] ) , so that each segment @xmath124 is identified with @xmath133 ( resp . , @xmath134 ) .
if @xmath63 and @xmath91 are adjacent on @xmath53 then the rectangular strips @xmath124 and @xmath135 are glued together at right angles as in figure [ crossing - fig ] .
a planar realization is _ fillable _ if it is possible to attach ( possibly non - convex ) polygons to the planar graph along closed cycles , so that the interior of the polygon does not include any endpoint of a line segment .
figure [ lines1-fig ] gives an example of a fillable planar realization , and figure [ lines2-fig ] gives an example of a non - fillable planar realization .
think of the planar realization as being embedded in @xmath5 .
let @xmath6 be the filled planar realization after gluing together each end of the horizontal strips to its opposite with a single positive full twist , and the end of each vertical strip to its opposite by a single negative full twist .
the boundary of @xmath6 is a link @xmath136 with distinguished seifert surface @xmath6 .
we call @xmath137 a
_ coxeter link _ associated to @xmath53 .
the link @xmath4 has an alternating planar diagram coming from drawing each vertical and horizontal hopf band as in figure [ leftright - fig ] . here
the shaded rectangle is the original neighborhood of the line segment associated to a vertex of @xmath53 .
the signs indicate over ( @xmath55 ) and under ( @xmath56 ) crossings .
thus we can see that for each vertex @xmath60 , when proceeding along @xmath124 there is always a @xmath56 sign on the right and a @xmath55 sign on the left , where @xmath56 indicates an upcoming underpass , and @xmath55 indicates an upcoming overpass .
since the signs are consistent on vertical and horizontal segments ( @xmath56 appears on the right and @xmath55 appears on the left no matter from which direction you approach an endpoint of a segment ) the link @xmath4 is alternating . since the surface @xmath6 can be obtained from a disk by hopf plumbings , the boundary of @xmath6 is a fibered link @xmath4 with fiber @xmath6 .
all the strips become annuli on @xmath6 .
the monodromy of the fibration is the product of right or left dehn twists around core curves of the annuli , right or left being determined by whether the twist is positive or negative @xcite .
let @xmath139 be the set of vertices of @xmath53 .
for @xmath60 , let @xmath140 be the closed curve defined by @xmath124 .
then the homology classes @xmath141 $ ] form a basis for @xmath142 , and the monodromy @xmath10 of @xmath6 is the product of positive dehn twists on @xmath140 for @xmath63 such that @xmath143 composed with the product of negative dehn twists on @xmath140 for @xmath63 such that @xmath144 .
let @xmath58 be the vector space of @xmath59-labelings of the vertices .
for @xmath60 , let @xmath61 $ ] be the corresponding element of @xmath58 giving the label @xmath62 on @xmath63 and @xmath64 on all other vertices of @xmath53 .
there is a commutative diagram @xmath145_{-c_{+- } } \ar[r ] & h_1(s;{\mathbf{r } } ) \ar[d]^{\phi_*}\\ & { \mathbf{r}}^{v_\gamma } \ar[r ] & h_1(s;{\mathbf{r } } ) } \ ] ] where the horizontal arrows taking @xmath61 $ ] to @xmath141 $ ] are isomorphisms .
the coxeter transformation decomposes as @xmath146 where @xmath147 , cf .
@xcite . by construction
, @xmath148 is also the seifert matrix for @xmath6 in @xmath149 with respect to the generators for homology given by the core curves of the attached hopf bands .
thus @xmath150 see e.g. @xcite , and is conjugate to @xmath138 . then @xmath154 \qquad c_- = \left [ \begin{array}{rrrrr } 1 & 0 & 0 & 0 & 0\\ 0 & 1 & 0 & 0 & 0\\ 0 & 0 & 1 & 0 & 0\\ -1 & -1 & 0 & -1 & 0\\ -1 & -1 & -1 & 0 & -1 \end{array } \right ] .\ ] ] setting the orientation on the seifert surface
@xmath6 so that the shaded area is oriented positively toward the viewer , we see that @xmath155 is the seifert matrix , and @xmath156 the coxeter transformation is given by @xmath157.\ ] ] the associated alexander and coxeter polynomials are : @xmath158 _ the link associated to a coxeter graph is not uniquely determined by the combinatorics of the graph .
figure [ tree2-fig ] shows two different planar embeddings of a coxeter graph .
the two links realizing these embeddings are distinct : one of them has an unknotted component , while the other does not . while for a large class of classical coxeter trees , two different planar embeddings always yield distinct but mutant links by a theorem of gerber @xcite , we do not know whether the same holds in the alternating - sign case .
_ in general , even if @xmath53 does not have a planar realization , it is possible to find a surface @xmath6 and a system of simple closed curves @xmath159 in one - to - one correspondence with @xmath139 such that ( see , e.g. @xcite ) . since @xmath53 is bipartite , the system of curves partitions into two multi - curves @xmath160 and @xmath161 that intersect transversally .
let @xmath162 and @xmath163 be the positive dehn twist along @xmath160 , respectively , the negative dehn twist along @xmath161 .
let @xmath164 .
we call @xmath14 a _ geometric realization _ of @xmath165 .
the proof follows along the same lines as the proof of proposition [ homdil - prop ] , the only difference being that the horizontal arrows in the commutative diagram need not be one - to - one or onto .
the cokernel is generated by boundary parallel curves whose homology classes are fixed by @xmath169 , hence their homology classes are contained in the eigenspace for @xmath62 .
let @xmath14 be a geometric realization of an alternating - sign coxeter graph @xmath165 .
then the eigenvalues of the homological action of @xmath10 are real and strictly positive by proposition [ bicolour - prop ] and lemma [ roots - lem ] .
this implies theorem [ realroots - thm ] .
similarly , theorem [ goldenmean - thm ] follows directly from proposition [ minimum - prop ] and lemma [ roots - lem ] .
[ interlacing - thm ] if @xmath170 and @xmath4 are alternating - sign coxeter links associated to @xmath90 and @xmath53 , respectively , where @xmath90 is a vertex extension of @xmath53 , then the roots of the alexander polynomial of @xmath170 and that of @xmath4 are interlacing .
99 n. acampo : _ sur les valeurs propres de la transformation de coxeter _ , invent .
* 33 * ( 1976 ) , no . 1 , 6167 .
a. e. brouwer , w. h. haemers : _ spectra of graphs _ , springer , new york , 2012 .
a. clay , d. rolfsen : _ ordered groups , eigenvalues , knots , surgery and l - spaces _ , math .
cambridge philos .
* 152 * ( 2012 ) , no . 1 , 115129 .
h. r. fox : _ some problems in knot theory _ ,
topology of 3-manifolds and related topics ( proc . the univ . of georgia institute ) ( 1962 ) , 168176 .
b. farb , d. margalit : _ a primer on mapping class groups _
, princeton university press ( 2011 ) .
d. gabai : _ the murasugi sum is a natural geometric operation _ , low dimensional topology ams cont . math
. studies * 20 * ( 1983 ) , 131144 .
y. gerber : _ positive tree - like mapping classes _ , phd thesis , university of basel ( 2006 ) .
m. hirasawa , k. murasugi : _ various stabilities of the alexander polynomials of knots and links _ , http://arxiv.org/abs/1307.1578 .
e. hironaka : _ chord diagrams and coxeter links _ , j. london math .
* 69 * ( 2004 ) , no . 2 , 243257 .
e. hironaka : _ mapping classes associated to mixed - sign coxeter graphs _ , http://arxiv.org/abs/1110.1013 .
r. howlett : _
coxeter groups and @xmath148-matrices _ , bull .
london math .
* 14 * ( 1982 ) , no . 2 , 137141 .
i. d. jong : _ alexander polynomials of alternating knots of genus two _
, osaka j. math .
* 46 * ( 2009 ) , no . 2 , 353371 . t. kanenobu : _ module dalexander des nuds fibrs et polynme de hosokawa des lacements fibrs _ , math .
notes kobe univ .
* 9 * ( 1981 ) , no . 1 , 7584 .
koseleff , d. pecker : _ on alexander - conway polynomials of two - bridge links _ , j. symbolic comput . * 68 * ( 2015 ) , part 2 , 215229 .
c. j. leininger : _ on groups generated by two positive multi - twists : teichmller curves and lehmer s number _ , geom .
* 8 * ( 2004 ) , 13011359 . l. lyubich , k. murasugi : _ on zeros of the alexander polynomial of an alternating knot _ , topology appl .
* 159 * ( 2012 ) , no . 1 , 290303 .
c. t. mcmullen : _ coxeter groups , salem numbers and the hilbert metric _ , publ .
hautes etudes sci .
* 95 * ( 2002 ) , 151183 .
k. murasugi : _ on the genus of the alternating knot .
i , ii _ , j. math .
japan * 104 * ( 1958 ) , 94105 , 235248 .
k. murasugi : _ on the alexander polynomial of alternating algebraic knots _ , j. austral .
soc . * 39 * ( 1985 ) , 317333 .
b. perron , d. rolfsen : _ on orderability of fibred knot groups _ , math .
cambridge philos .
* 135 * ( 2003 ) , no . 1 , 147153 . b. perron , d. rolfsen : _ invariant ordering of surface groups and 3-manifolds which fibre over @xmath171 _ , math
cambridge philos .
soc . * 141 * ( 2006 ) , no . 2 , 273280 .
j. f. mckee , p. rowlinson , and c. j. smyth : _ salem numbers and pisot numbers from stars _ , number theory in progress , vol .
1 ( zakopane - kocielisko , 1997 ) , de gruyter ( 1999 ) , 309319 .
p. ozsvth , z. szab : _ heegaard floer homology and alternating knots _ , geom .
* 7 * ( 2003 ) , 225254 .
j. rasmussen : _ khovanov homology and the slice genus _ , invent . math . * 182 * , no . 2 , 419447 .
d. rolfsen : _ knots and links _ ,
chelsea publishing ( 1976 ) .
j. stallings : _ construction of fibred knots and links _ , algebraic and geometric topology , proc .
pure math . * 32 * ( 1978 ) , 5560 , amer . math .
soc . , providence , r.i .
r. steinberg : _ finite reflection groups _ , trans .
soc . * 91 * ( 1959 ) , 493504 .
a. stoimenow : _ hoste s conjecture and roots of link polynomials _ ,
preprint 2013 , http://stoimenov.net/stoimeno/homepage/abstracts.html .
w. thurston : _ on the geometry and dynamics of diffeomorphisms of surfaces _ , bull .
( n.s . )
* 19 * ( 1988 ) , no . 2 , 417431 . c. g. wagner : _ newton s inequality and a test for imaginary roots _ , two - year college math .
j. * 8 * ( 1977 ) , no . 3 , 145147 . | alternating - sign hopf plumbing along a tree yields fibered alternating links whose homological monodromy is , up to a sign , conjugate to some alternating - sign coxeter transformation . exploiting this tie ,
we obtain results about the location of zeros of the alexander polynomial of the fibered link complement implying a strong case of hoste s conjecture , the trapezoidal conjecture , bi - orderability of the link group , and a sharp lower bound for the homological dilatation of the monodromy of the fibration .
the results extend to more general hyperbolic fibered 3-manifolds associated to alternating - sign coxeter graphs . |
the way shell closures and single - particle energies evolve as functions of the number of nucleons is presently one of the greatest challenges to our understanding of the basic features of nuclei .
doubly - magic nuclei are particularly important and closed shell nuclei like @xmath0ni and @xmath5sn have been the focus of several experiments during the last years @xcite .
their structure provides important information on theoretical interpretations and our basic understanding of matter .
in particular , recent experiments @xcite have aimed at extracting information about single - particle degrees of freedom in the vicinity of @xmath0ni .
experimental information from single - nucleon transfer reactions and magnetic moments @xcite , can be used to extract and interpret complicated many - body wave functions in terms effective single - particle degrees of freedom .
transfer reactions provide for example information about the angular distributions , the excitation energies and the spectroscopic factors of possible single - particles states .
if one can infer from experimental data that a single - particle picture is a viable starting point for interpreting a closed - shell nucleus like @xmath0ni , one can use this nucleus as a basis for constructing valence - space effective interactions .
these interactions can in turn be used in shell - model calculations of nuclei with several valence nucleons above the @xmath6 and @xmath7 filled shells of @xmath0ni .
recent measurements of spectroscopic factors of @xmath4ni in high - energy knockout reactions @xcite seem to indicate that low - lying states in @xmath4ni can be characterized as single - particle states on top of @xmath0ni as a closed - core nucleus .
large - scale shell - model calculations by horoi _
@xcite corroborate these findings , whereas a recent experiment on magnetic moments of the ground state of @xmath4cu @xcite , expected to be described as one valence proton outside @xmath0ni , resulted in much smaller moments than those expected from a single - particle picture .
similarly , large transition matrix elements between the @xmath8 ground state and the first excited @xmath9 state indicate that the @xmath0ni core is rather soft @xcite , or stated differently , it implies a rather fragmented single - particle picture . on the other hand
, one ought keep in mind that quenchings of spectroscopic factors to about 60% are common even for good closed shell nuclei @xcite .
experimentally , spectroscopic factors are defined as the ratio of the observed reaction rate with respect to the same rate calculated assuming a full occupation of the relevant single - particle states .
they are therefore often interpreted as a measure of the occupancy of a specific single - particle state . however , from a strict theoretical point of view spectroscopic factors are not occupation numbers but a measure of what fraction of the full wave function can be factorized into a correlated state ( often chosen to be a given closed - shell core ) and an independent single - particle or single - hole state .
large deviations from the values predicted by an independent - particle model , point to a strongly correlated system . in this regime ,
collective excitations that behave like single - particle degrees of freedom that is _ quasiparticles_can still arise .
the above mentioned large - scale shell - model calculations @xcite have been performed in one major shell , the @xmath2-shell , with an effective interaction fitted to reproduce properties of several nuclei that can be interpreted in terms of these single - particle states .
the number of possible slater determinants that can be constructed when distributing eight valence protons and eight valence nucleons in the @xmath2 shell is more than @xmath10 .
this means that the inclusion of more complicated particle - hole excitations from shells below and above the @xmath2 shell , are well beyond present capabilities of large - scale diagonalization methods @xcite .
the hope is that an effective interaction tailored to one major shell includes as many as possible of these neglected particle - hole excitations .
however , there are other many - body methods that allow for a computational scheme which accounts for a systematic inclusion of more complicated many - body corrections
. typical examples of such many - body methods are coupled - cluster methods @xcite , various types of monte carlo methods @xcite , perturbative many - body expansions @xcite , self - consistent green s functions ( scgf ) methods @xcite , the density - matrix renormalization group @xcite and _ ab initio _ density functional theory @xcite , just to mention some of the available methods . the green s function monte carlo method @xcite and the no - core shell - model approach @xcite have been successfully applied to the theoretical description of light nuclei with mass numbers @xmath11 , and hamiltonians based on nucleon - nucleon and three - nucleon interactions . however
, present experimental studies of nuclear stabilities are now being pushed to larger mass regions , with mass numbers from @xmath12 to @xmath13 .
traditionally , this has been the realm of the nuclear shell - model and nuclear density - functional theory .
these methods employ hamiltonians and density functionals with phenomenological corrections and are not directly related to the vacuum nucleon - nucleon interaction employed in _ ab initio _ calculations ( exceptions are found when perturbative many - body methods are used @xcite ) .
however , in selected medium - mass nuclei , _ ab initio _ structure calculations can be performed using approaches like coupled - cluster and green s functions theories .
these methods allow studying ground- and excited - state properties of systems with dimensionalities beyond the capability of present large - scale diagonalization approaches , with a much smaller numerical effort when compared to diagonalization methods aiming at similar accuracies .
the accuracy of these methods is sufficiently high to attribute an eventual disagreement between experimental data and theoretical results to missing physics in the hamiltonian . in this way
, such calculations help to increase our understanding of the nuclear interaction on a very fundamental level .
recent coupled - cluster calculations @xcite have reported practically converged results of the ground state of medium - mass nuclei like @xmath14ca , @xmath15ni and @xmath15ca using the bare chiral interaction n@xmath1lo @xcite .
these calculations employed a harmonic oscillator basis to construct the single - particle basis and included correlations of the so - called singles and doubles types .
it means that one - particle - one - hole and two - particle - two - hole correlations acting on a many - body slater determinant were summed to infinite order .
recently , results with three - particle - three - hole correlations have also been obtained @xcite .
the calculations were performed in a harmonic oscillator basis containing up to fifteen major shells and resulted in basically converged ground state properties for a given hamiltonian .
another method with a strong potential for performing _ ab initio _ calculations of nuclei beyond @xmath16 is self - consistent green s functions theory . differently from the coupled - cluster and no - core shell model ,
this method does not construct the wave function but evaluates directly the energies and transition matrix elements for the transfer of one or more nucleons .
another important point is that the self - energy
the central component of the formalism has been shown to be an exact optical potential .
on the one hand , one can employ the formalism for pure _ ab initio _ studies .
on the other , the strong link with the response to experimental probes can be used to constrain and improve phenomenological models .
an example of this approach is the dispersive optical model recently derived for chain of calcium isotopes @xcite .
this global optical potential reproduces with high accuracy the known elastic scattering data , up to energies of 200 mev .
thus , green s functions hold a promise of _ both _ bridging nuclear structure and reactions and for connecting the ( relatively few ) isotopes amenable of _ ab initio _ calculations to the rest of the nuclear landscape . in practical applications of green s functions theory ,
one expands the self - energy in terms of resummations of feynman diagrams and truncates the series in a way that allows for further systematic improvements of the formalism .
a powerful scheme for non - perturbative expansions is the so called faddeev random phase approximation ( frpa ) that explicitly accounts for particle - vibration couplings @xcite .
first applications of this approach were devoted to @xmath17o .
one single calculation yielded the basic information to be used for microscopic studies of spectroscopic factors @xcite , the excitation spectrum @xcite , two - nucleon emission @xcite and the nucleon - nucleus optical potential @xcite .
thus , the frpa method pursues a global description of the many - body dynamics , far beyond ground state properties alone .
another difference with the coupled - cluster approach is that green s functions allow , via a diagrammatic approach , to directly introduce the correlations outside the model space that are associated with short - range degrees of freedom @xcite .
it can therefore be applied to interactions with strong short - range cores ( for example , the argonne model @xcite was employed in ref .
@xcite ) .
self - consistent frpa calculations of @xmath17o were first performed by the authors of ref .
@xcite , and subsequently extended to fully _ ab initio _ calculations in spaces up to eight oscillator shells @xcite . to our knowledge , the combination of the random phase approximation phonons and the proper treatment of the energy dependence of the interaction vertex , makes this the most accurate evaluation of single - particle states available for this nucleus . in this work
we extend the range of applications of this formalism to studies of quasiparticle states around @xmath0ni , as a first application of faddeev random phase approximation to @xmath2 shell nuclei .
as mentioned above , there is quite some experimental interest in single - particle properties around @xmath0ni .
this work is organized as follows .
[ formalism ] reviews the faddeev random phase approximation approach and discusses the approximations made to calculate the self - consistent propagator .
the convergence of single - particle properties is discussed in sec .
[ calc ] and the results for the spectral function are described in sec .
[ res_spctfnct ] .
we refer the reader directly interested to discussion of physics results to the latter section .
conclusions are drawn in sec .
[ concl ] .
this section serves as an overview of the formalism we employ . the treatment of short - range physics and the implementation of self - consistency , which are improved with respect to previous works ,
are discussed in details .
the faddeev random phase approximation expansion is also introduced , but we refer the reader for more details in refs .
@xcite . in the framework of green s function theory ,
the object of interest is the single - particle propagator , instead of the many - body wave function . in the following
, greek indices @xmath18 label the orthonormal basis set of single - particle states included in the model space , while latin indices @xmath19 refer to many - body states .
we employ the convention of summing over repeated indices , unless specified otherwise .
the single - particle propagator can be written in the so - called lehmann representation as @xcite @xmath20 where @xmath21 ( @xmath22 ) are the spectroscopic amplitudes , @xmath23 ( @xmath24 ) are the second quantization annihilation ( creation ) operators and @xmath25 ( @xmath26 ) . with these definitions ,
@xmath27 and @xmath28 are the eigenstates , while @xmath29 and @xmath30 are the corresponding energies of the ( @xmath31)-nucleon system .
therefore , the poles of the single - particle propagator reflect the energy transfer observed in pickup and knockout reactions .
the single - particle propagator @xmath32 enters the dyson equation as @xmath33 it depends on the irreducible self - energy @xmath34 .
the latter can be written as the sum of two terms .
the first terms describes the average mean - field ( mf ) while the second term contains dynamic correlations , @xmath35 in eq .
( [ eq : dys ] ) , @xmath36 is the so - called unperturbed single - propagator , corresponding to nucleons moving under the effect of the kinetic energy part of the total hamiltonian .
the localization of the single - particle states in the nuclear mean - field is due to the term @xmath37 , which extends the hartree - fock potential to that of a fully correlated density matrix .
the term @xmath38 represents the antisymmetrized matrix elements of the nucleon - nucleon interaction . in this work ,
these will be approximated by an effective interaction , eq .
( [ eq : veff ] ) , discussed in sec . [ form_bhf ] . eq .
( [ eq : sigma1 ] ) introduces the two - particle - one - hole ( _ 2p1h _ ) and two - hole - one - particle ( _ 2h1p _ ) irreducible propagator @xmath39 . in its feynman expansion
, this propagator contains all diagrams with any number of particle and hole lines _ except _ those that allow the intermediate propagation of one single line .
it can therefore be interpreted as carrying the complete information on all configurations that can not be reduced to a nucleon interacting with the average nuclear field . in particular
, it includes the coupling of single - particle states to collective vibrations like giant resonances .
the theoretical spectroscopic factors @xmath40 and @xmath41 for removal and addition , respectively , of a nucleon , are given by the normalization integral of the corresponding overlap wave functions .
in the notation of eq .
( [ eq : g1 ] ) , these are @xmath42 the hole states are normalized according to @xmath43 which follows directly from the dyson equation ( [ eq : dys ] ) . the same relation applies to particle states , with @xmath44 replaced by @xmath45 . because of the analytical properties of @xmath34 , the derivative term results always in a negative contribution , leading thereby to a quenching of the spectroscopic factors .
it must be stressed that eqs .
( [ eq : dys ] ) and ( [ eq : sigma1 ] ) do not involve any approximation .
therefore , the full knowledge of @xmath37 and @xmath39 would be equivalent to the exact solution of the schrdinger equation .
in practical calculations , it is always necessary to truncate the full space of available single - particle states to a finite model space and to select a limited set of many - body correlations .
the approximations employed in the present work to evaluate these quantities are discussed in the rest of this section .
the present calculations were performed within a large but finite set of harmonic oscillator states , including single - particle states up to ten major shells . in order to treat the short - range part of the nucleon - nucleon interaction ,
one must resum explicitly the series of ladder diagrams for two nucleons outside this model space .
these contributions are included in the self - energy in two different ways .
firstly , they are explicitly added to the mean - field part , @xmath37 , in order to reconstruct the contribution of short - range correlations in the full hilbert space .
secondly , they are included in a regularized effective hamiltonian which is used to calculate the long - range part of correlations described by @xmath39inside the chosen model space .
this approach leads to calculating the well known @xmath46-matrix @xcite , which is then used as an energy dependent effective interaction inside the model space in which _ all _ states belonging to the model space are pauli blocked since diagrams among them are already included in the many - body calculation of @xmath39 . ] . in this case
the @xmath37 part of the self - energy will also depend on energy .
this is given by @xmath47 where @xmath48 are the matrix elements of the @xmath46-matrix interaction .
( [ eq : h_bhf ] ) differs from the standard brueckner - hartree - fock potential in the fact that the mean - field is not represented by a set of independent nucleons filling various orbits .
rather , the medium is described by the hole spectroscopic amplitudes @xmath49 that lead to the fully correlated density matrix .
since the latter are obtained by solving the dyson equation , they must be obtained iteratively in a self - consistent way . whenever the second term in eq .
( [ eq : sigma1 ] ) is neglected , this procedure simply reduces to solving the standard brueckner - hartree - fock equations .
however , as soon as the response of the medium @xmath39 is accounted for , the single - particle propagator becomes fragmented and @xmath37 describes the interaction of a particle with the `` correlated '' medium .
we remind that using the @xmath46-matrix in eq .
( [ eq : h_bhf ] ) corresponds to summing the mean - field term form eq .
( [ eq : sigma1 ] ) and ladder diagrams with intermediate two - particle states outside the chosen model space .
this partitioning procedure has two main consequences .
first , the effects of short - range physics at the two - body level on the total energy are included in the renormalized interaction .
this leads to a softer force that can be applied within a `` low - momentum '' model space . secondly , due to the explicit energy dependence
, the term @xmath50 contributes as well to the normalization of spectroscopic factors , eq .
( [ eq : sf ] ) .
this provides a natural way to determine the amount of strength that the free interaction would admix into configurations outside the model space , see for example the discussion of refs .
thus , the present approach differs from methods based on the renormalization group , where instead an explicit renormalization of the effective operators would be required @xcite .
the energy dependence of the @xmath46-matrix becomes cumbersome in calculating the polarization propagator .
we define therefore a static effective interaction for our model space to be used in our calculations of the second term on the right - hand side of eq .
( [ eq : sigma1 ] ) . to do this we evaluate the average energy for the _ harmonic oscillator _ single - particle states according to @xmath51 where the sum is limited to those states that correspond to filled orbits in the independent particle model .
note that in eq .
( [ ho_spe ] ) the single - particle energies are derived iteratively while the oscillator wave functions remain unchanged . clearly , these orbits are a crude approximation to the real quasiparticle states and will not be used to construct @xmath39 in sec .
[ form_frpa]which will be rather determined in a self - consistent fashion .
the main purpose of the above procedure is to yield a prescription for obtaining a starting energy independent effective interaction . following gad and mther @xcite
, we use the single - particle energies from eq .
( [ ho_spe ] ) to define an interaction for the given model space by @xmath52 \ ; .
\label{eq : veff}\ ] ] the @xmath46-matrix can be computed according ref .
@xcite for negative energies , up to about @xmath53 mev . for larger values
we fix the starting energy to @xmath54 mev .
note that the starting energy appearing in eq .
( [ eq : h_bhf ] ) is shifted by the quasihole poles @xmath55 .
this ensures that the energy dependence of @xmath50 is fully accounted for all quasiparticle states in the @xmath2 shell .
the polarization propagator @xmath39 can be expanded in terms of simpler green s functions that involve the propagation of one or more quasiparticle states .
this approach has the advantage that it aids in identifying key physics ingredients of the many - body dynamics . by truncating the expansion to a particular subsets of diagrams or many - body correlations
, one can then construct suitable approximations to the self - energy .
moreover , since infinite sets of linked diagrams are summed the approach is non - perturbative and satisfies the extensivity condition @xcite .
this expansion also serves as a guideline for systematic improvements of the method .
following refs .
@xcite , we first consider the particle - hole polarization propagator that describes excited states in the @xmath56-particle system @xmath57 and the two - particle propagator that describes the addition / removal of two particles @xmath58 these green s functions contain in their lehmann representations all the relevant information regarding the excitation of particle - hole and two - particle or two - hole collective modes . in this work
we are interested in studying the influence of giant resonance vibrations , which can be described within the random phase approximation ( rpa ) . in the faddeev rpa approach , the propagators of eqs .
( [ eq : pi ] ) and ( [ eq : g2 ] ) are then evaluated by solving the usual rpa equations , which are depicted diagrammatically in fig .
[ fig : rpaeq ] .
since these equations reflect two - body correlations , they still have to be coupled to an additional single - particle propagator , as in fig .
[ fig : faddex ] , to obtain the corresponding approximation for the two - particle - one - hole and two - hole - one - particle components of @xmath39 .
this is achieved by solving two separate sets of faddeev equations , as discussed in ref .
@xcite . ) .
the full lines represent the independent - particle model propagator @xmath59 , which is employed instead of the fully dressed one .
see the text for details . ]
taking the two - particle - one - hole ( 2p1h ) case as an example , one can split @xmath60 in three different components @xmath61 ( @xmath62 ) that differ from each other by the last pair of lines that interact in their diagrammatic expansion , @xmath63 + \sum_{i=1,2,3 } \bar{r}^{(i)}_{\alpha \beta \gamma ,
\mu \nu \lambda}(\omega ) \ ; , \label{eq : faddfullr } % \end{eqnarray}\ ] ] where @xmath64 is the _ 2p1h _ propagator for three freely propagating lines .
these components are solutions of the following set of faddeev equations @xcite @xmath65 \ ; , ~ ~ i=1,2,3 ~ % \right ] \ ; , \nonumber\end{aligned}\ ] ] where ( @xmath66 ) are cyclic permutations of ( @xmath67 ) . the interaction vertices @xmath68 contain the couplings of a particle - hole ( _ ph _ ) , see eq .
( [ eq : pi ] ) , or two - particle / two - hole ( _ pp / hh _ ) , see eq .
( [ eq : g2 ] ) , collective excitations and a freely propagating line .
the propagator @xmath39 which we employ in eq .
( [ eq : sigma1 ] ) is finally obtained by @xmath69 where the matrix @xmath70 has the effect of renormalizing the strength of the dynamic self - energy .
this correction ensures consistency with perturbation theory up to third order .
the explicit formulae of the matrices @xmath68 and @xmath70 are given in terms of the propagators of eqs .
( [ eq : pi ] ) , ( [ eq : g2 ] ) and ( [ eq : gspref ] ) , and the interaction @xmath38 .
they are discussed in detail in ref .
the calculation of the _
2h1p _ component of @xmath39 follows completely analogous steps . ) ( left ) .
the corresponding contribution to the self - energy , obtained upon insertion into eq .
( [ eq : sigma1 ] ) , is also shown ( right ) . ]
the present formalism includes the effects of _ ph _ and _ pp / hh _ motion simultaneously , while allowing interferences between these modes .
these excitations are evaluated here at the rpa level and are then coupled to each other by solving eqs .
( [ eq : faddtda ] ) .
this generates diagrams as the one displayed in fig .
[ fig : faddex ] .
the faddeev equations also ensure that the pauli principle is correctly taken into account at the _
2p1h _ and _ 2h1p _ level .
in addition , one can in principle employ dressed single - particle propagators in these equations to generate a fully self - consistent solution , as done in refs .
@xcite for valence orbits around @xmath17o . in the self - consistent green s function approach , both the @xmath37 part of the self - energy and the polarization propagator @xmath39 are expressed directly in terms of the exact single - particle propagator @xmath71 .
the lines in figs .
[ fig : rpaeq ] and [ fig : faddex ] should thus represent the fully dressed propagator obtained by solving the dyson equation .
since the degrees of freedom contained in eq .
( [ eq : g1 ] ) are excitations of the fully correlated system , the formalism does not depend on an explicit reference state .
normally , one first computes eq .
( [ eq : sigma1 ] ) in terms of an approximate propagator . the solution of eq .
( [ eq : dys ] ) is then used to calculate an improved self - energy and the procedure is iterated to convergence .
baym and kadanoff have shown that the self - consistency requirement implies the conservation of both microscopic and macroscopic properties @xcite .
intuitively , the self - consistency requirement becomes important whenever dynamical correlations modify substantially the response with respect to the hartree - fock mean - field ( an example is the band - gap error problem in diamond crystals @xcite ) .
when applying standard hartree - fock theory to nuclear structure , most realistic interactions predict unbound nuclei and valence orbits in the continuum .
this is a very poor starting point for any application of perturbation theory and other many - body techniques .
however , the self - consistent approach requires using correlated quasiparticle energies and wave functions [ the poles and residues of eq .
( [ eq : g1 ] ) ] .
these degrees of freedom form an optimal starting point for studies of many - body dynamics at the fermi surface .
accounting for the fragmentation of the single - particle propagator in the faddeev random phase approximation increases the computational load as one moves to larger nuclei and model spaces .
in this situation it is convenient to expand @xmath39 in terms of an independent - particle model ( ipm ) propagator .
this should approximate the dressed one but with a limited number of poles .
thus , we solve eq .
( [ eq : faddtda ] ) in terms of @xmath72 where @xmath73 represent the set of occupied orbits .
the single - particle energies @xmath74 and wave functions @xmath75 are chosen such that @xmath59 coincides with the real propagator @xmath71 at the fermi surface . to do this
we define the following moments of the poles of eq .
( [ eq : g1 ] ) .
@xmath76^p } + \sum_k \frac { { \cal y}^{k}_{\alpha } \ ; \left ( { \cal y}^{k}_{\beta } \right)^ * } { \left [ e_f - \varepsilon^{-}_k \right]^p } \ ; , ~ ~ p=0,1,2,\ldots \label{eq : mi}\ ] ] where @xmath77 is the fermi energy .
( [ eq : gspref ] ) is determined by imposing @xmath78 and @xmath79 .
the purpose of eq .
( [ eq : mi ] ) is to define a set of effective single - particle orbits and energies that conserve the total spectroscopic strength carried by the self - consistent propagator and the centroids of its fragmented states .
while effective single - particle properties form an appropriate starting point to evaluate @xmath39 , it remains clear that they only represent average quantities .
instead , it is eq .
( [ eq : g1 ] ) that must be related to experiment .
the propagator @xmath59 is derived from eq .
( [ eq : g1 ] ) and it still needs to be evaluated in a iterative way .
therefore , the resulting propagator @xmath39 is ( partially ) self - consistent .
we stress that eq .
( [ eq : h_bhf ] ) can be calculated easily from the fully dressed propagator .
thus self - consistency is achieved _ exactly _ at the mean - field level .
the present calculations were performed using a harmonic oscillator basis and including up to ten major harmonic oscillator shells .
we label these spaces with @xmath80 , 5 , 7 , or 9 , where @xmath81 . for the largest model space employed , @xmath82 , all partial waves with orbital angular momentum @xmath83
were included . this amount to 368 single - particle states for each particle species , protons and neutrons in our case .
the total number of available slater determinants for @xmath0ni without any particular restrictions is proportional to the product of the two binomials @xmath84 a number which clearly exceeds the capabilities of any direct large - scale diagonalization procedure .
the codes utilize a @xmath85-coupling scheme to decouple the faddeev equations of eq .
( [ eq : faddtda ] ) . at each iteration ,
the rpa equations are solved in the particle - hole and the particle - particle and hole - hole channels using the single - particle orbits and energies from eq .
( [ eq : gspref ] ) .
the resulting propagators are inserted in eqs .
( [ eq : faddtda ] ) , which in turn are casted into a non hermitian eigenvalue problem @xcite . in our largest calculation
we diagonalize dense matrices of dimension up to @xmath86 two - particle - one - hole states .
this number is bound to increase when larger nuclei are investigated or more details of nuclear fragmentation ( that is more poles ) are included in @xmath59 ( [ eq : gspref ] ) .
the numerical implementation of the faddeev random phase approximation required careful optimization in evaluating the elements of the faddeev matrix and a proper generalization of the arnoldi algorithm @xcite to employ multiple pivots .
similar improvements allowed to extend large scale calculations from the @xmath87 mass region to the @xmath88 mass region .
the dimensions reached in this work represent roughly the upper limit when using table top single processor computers .
obviously , there is much to gain by taking advantage of modern supercomputer facilities and future research efforts should be put into parallelization of the present algorithms .
model spaces of eight to ten major shells are large enough for a proper description of the response due to long - range correlations .
these include excitations of several mevs into the region of giant resonances .
the effects of short - range physics are also included by using a @xmath46-matrix and an effective interaction as discussed in sec . [ form_bhf ] .
these are derived using the chiral nucleon - nucleon interaction n@xmath1lo by entem and machleidt @xcite .
this interaction employs a cutoff of @xmath89 mev .
typical realistic two - nucleon interactions fail in reproducing the spin - orbit splittings and gaps between different shells . in particular , for the @xmath6 and @xmath7 subshell closures these lead to an underestimation of the gap at the fermi surface .
in these cases , a complete diagonalization of the hamiltonian would predict a deformed ground state of these nuclei even when they are experimentally known to be good spherical closed - shells systems @xcite .
this issue can be cured with a simple modification of the monopole strengths of the interaction .
recently , zuker has reported that the same correction works well for several isotopes throughout the nuclear chart and proposed that this may be interpreted as a signature of missing three - nucleon interactions @xcite .
the inclusion of three - nucleon interactions to the faddeev random phase approximation formalism is beyond the scope of the present work .
however , it will be shown in sec .
[ calc_kdep ] that properly reproducing the fermi gap is crucial in order to obtain meaningful results for the valence space spectroscopic factors .
thus , we follow ref .
@xcite and modify the monopoles in the n@xmath1lo interaction model as @xmath90 where @xmath91 and @xmath92 stand for the brueckner - hartree - fock states associated to the @xmath93 and the ( @xmath94,@xmath95,@xmath96 ) orbits , respectively . in the limit of large spaces , the brueckner - hartree - fock orbits converge to the hartree - fock states and this correction becomes independent of the choice of the single - particle basis .
note that the prescription of eq .
( [ eq : monopole ] ) modifies only a few crucial matrix elements while about six millions of them are defined in the @xmath82 model space .
we also note that the present truncation of the model space , in terms of the number of oscillator shells , does not separate exactly the center of mass motion .
coupled - cluster calculations have shown that the error introduced by this truncation becomes negligibly small for large model spaces such as the ones employed here and therefore it does not represent a major issue @xcite . in calculations of binding energies , it customary to subtract the operator for the kinetic energy of the center of mass directly from the hamiltonian .
this term automatically corrects for the zero point motion in oscillator basis but it depends explicitly on the number of particles . in this work ,
we are interested in transitions to states with different numbers of nucleons [ ( a@xmath971 ] and aim at computing directly the differences between the total energies .
therefore , the above correction should _ not _ be employed in the present case .
one may note that the separation of the center - of - mass motion is an issue related to the choice made for the model space , rather than the many - body method itself .
for example , expressing the propagators directly in momentum space would allow an exact separation .
in this situation , the transformation between the center - of - mass and laboratory frames for systems with a nucleon plus a @xmath56 nucleons ( or @xmath98 nucleons ) core would also be simple . and the @xmath93 valence orbits with respect the ph gap @xmath100 . for each model space ,
different points correspond to different choices of @xmath99 in the range @xmath101 mev . ]
( [ eq : monopole ] ) introduces a single parameter ( @xmath99 ) in our calculations .
the reason for this modification is that the spectroscopic factors of the valence orbits are strongly sensitive to the particle - hole gap .
this sensitivity is to be expected since collective modes in the @xmath0ni core are dominated by excitations across the fermi surface .
smaller gaps imply lower excitation energies and higher probability of admixture with valence orbits . in order to extract meaningful predictions for spectroscopic factors
it is therefore necessary to constrain the fermi gaps for protons and neutrons to their experimental values . to investigate this dependency we repeated our calculations for values of @xmath99 in the range @xmath101 mev .
[ sf_vs_k ] shows the resulting neutron spectroscopic factors for the valence @xmath102 quasiparticle and @xmath103 quasihole .
these are plotted as a function of the calculated particle - hole gap @xmath104 .
the results correspond to model spaces of different dimensions ( eight or ten oscillator shells ) and oscillator frequencies ( @xmath105 or 18 mev ) .
the gap @xmath100 increases with @xmath99 but the dependence on the model space is weak .
we notice that , once the experimental value of @xmath100 is reproduced , the spectroscopic factors are well defined and found to be converged with respect to the given model space .
all results reported below were obtained with a fixed value of @xmath106 mev . in the @xmath82 model space and an oscillator energy @xmath105
mev , this choice reproduces the experimental gaps at the fermi surface for both protons and neutrons to an error within @xmath107 kev . from fig .
[ sf_vs_k ] one infers that the calculated spectroscopic factors are reliable to within @xmath108% of the independent - particle model value .
particle energy and the @xmath93 hole energy with respect to the oscillator frequency and the size of the model space . ]
corresponding to the valence @xmath2 orbits .
calculations were performed for @xmath82 and @xmath83 . ] fig .
[ spe_vs_msp ] shows the dependence of the neutron @xmath94 particle and the @xmath93 hole energies with respect to the oscillator frequency and the size of the model space . as can be seen from this figure , the single - particle energies for these two single - particle states tend to stabilize around eight to ten major shells
. this finding concords both with coupled - cluster calculations that employ a @xmath46-matrix as effective interaction for @xmath17o , see refs .
@xcite , and with analogous green s functions studies @xcite .
it remains however to make an extensive comparison between coupled - cluster theory and the green s function approach in order to find an optimal size of the model space with a given nucleon - nucleon interaction .
finally , we plot in fig .
[ spe_vs_hw ] the neutron valence single - particle energies for all the single - particle states in @xmath2-shell .
the latter results were obtained with our largest model space , ten major shells with @xmath82 and the single - particle orbital momentum @xmath83 . as can be seen from this figure
, there is still , although weak , a dependence upon the oscillator parameter . to perform calculations beyond ten major shells
will require non trivial extensions of our codes .
lcccccc & & & & + & & & & + & & frpa & exp . &
& frpa & exp .
+ @xmath4ni : + @xmath109 & & -11.43 & -9.134 & & 0.63 & + @xmath110 & & -10.80 & -9.478 & & 0.59 & + @xmath111 & & -12.78 & -10.247 & & 0.65 & 0.58(11 ) + @xmath112ni : + @xmath113 & & -19.22 & -16.641 & & 0.72 & + + @xmath4cu : + @xmath114 & & -1.28 & + 0.417 & & 0.66 & + @xmath115 & & -0.58 & & & 0.60 & + @xmath116 & & -2.54 & -0.695 & & 0.67 & + @xmath112co : + @xmath117 & & -9.08 & -7.165 & & 0.73 & + our results for spectroscopic factors for the @xmath2-shell valence orbits and the corresponding single - particle energies are collected in table [ tab : sf ] . in general , the modified n@xmath1lo interaction predicts single - particle energies about @xmath118 mev lower than the experimental ones . the coulomb shift between corresponding neutron and proton orbits is calculated to be about @xmath119 mev and it is closer to the empirical value of @xmath120 mev . for the oscillator parameter chosen , @xmath105 mev , we obtain an inversion of the @xmath95 and @xmath96 excited states in @xmath4ni , with respect to the experiment .
however , this discrepancy disappears for larger values of @xmath121 ( see fig .
[ spe_vs_hw ] ) .
this effect is in fact smaller than the residual dependence on the model space and therefore no conclusion can be made about the ordering for the fully converged result .
the spectroscopic factor for the transition between the ground states of @xmath4ni and @xmath0ni was extracted from high energy knockout reactions in ref .
the self - consistent faddeev random phase approximation result for this quantity yield 65% of the independent - particle model value , and agrees with the empirical data within experimental uncertainties .
the theoretical spectroscopic factors for the excited states in @xmath4ni are similar , with the @xmath96 state being somewhat smaller , at about 59% .
a larger value is obtained for knockout to the ground state of @xmath112ni , which is predicted to be 72% . the results for proton transfer to particle ( hole ) states in @xmath4cu ( @xmath112co )
are only slightly larger . according to the analysis of fig .
[ sf_vs_k ] , it is expected that these predictions are converged within 1 - 2% of the independent - particle model values .
past studies @xcite have questioned whether low - energy quasiparticle states in @xmath4ni are strongly admixed to excitations of a soft @xmath0ni core .
the results obtained here do not suggest substantial differences with respect to other known closed shell nuclei .
the spectroscopic factors from table [ tab : sf ] are in line with observations from stable nuclei @xcite and support the hypothesis that @xmath0ni is a good closed - shell nucleus . in our calculations
we find that the @xmath94 quasiparticle state of @xmath4ni carries 65% of the strength for this orbit .
another 20% is located in the particle region below @xmath122=2 mev ( above this energy strength associated with the @xmath123 shell starts to appear ) , and about 3% is in the hole region above @xmath124 mev ( see fig .
[ fig : spctfnct ] ) .
similarly , the @xmath93 state has 72% of the independent - particle model strength in the quasihole peak ( the ground state of @xmath112ni ) , 10% in the fragmented hole region , and 3% in the fragmented particle region .
this analysis confirms that the main mechanism responsible for the quenching of the spectroscopic factors lies in the admixture between single - particle states and collective excitations in the region of giant resonances @xcite . due to these correlations a large part of the missing strength from the valence peak
is shifted and spread over an adjacent region about @xmath125 mev wide .
further reduction of the spectroscopic factors comes from the mixing with configurations at much higher energies and momenta and is accounted for through the energy dependence of eq .
( [ eq : h_bhf ] ) .
the information carried by the calculated single - particle propagator @xmath71 is collected in fig .
[ fig : spctfnct ] for neutrons and partial waves up to @xmath126=3 .
the plot shows the spectral strength associated with each pole of @xmath71 .
fragments below the fermi surface ( @xmath77 ) refer to the separation of a neutron ( @xmath112ni ) , while those above correspond to neutron addition ( @xmath4ni ) .
the poles corresponding to the @xmath2 valence orbits are indicated by arrows .
these are the same single - particle energies that have been discussed in sec .
[ calc_mspconv ] and tab .
[ tab : sf ] . as noted above
, the fragments found at slightly higher energies ( just above @xmath127 mev ) originate from the mixing of these orbits with two - particle - one - hole configurations and collective excitations of the nucleus .
the overall fragmentation effect is substantial but not strong enough to destroy the single - particle character of the principal quasiparticle peaks ( note that a logarithmic scale has been chosen in fig .
[ fig : spctfnct ] in order to make smaller fragments of the spectral distribution more visible ) .
this observation supports the use of valence single - particle states as the relevant degrees of freedom that govern low - lying excitations , as assumed in conventional shell - model applications .
we note that the question whether a system can be approximated as a good shell closure is better addressed by analyzing the spectroscopic factors and strength distribution rather than occupation numbers , since the latter are integrated quantities ) . especially for deeply bound orbits ,
it is possible that a strong fragmentation pattern still leads to large occupation numbers . ] .
while unoccupied states can be probed by the addition of a nucleon , occupied states are accessed by knockout to states of the @xmath128 nucleon system .
a similar fragmentation pattern is therefore seen for the @xmath93 orbit but reversed below the fermi surface .
interestingly enough , the faddeev random phase approximation predicts that states corresponding to orbits in the @xmath129 and @xmath130 shells maintain a strong single - particle character even though they are further apart from the fermi surface .
the fragmentation of these orbits requires excitations across shells of different parity ( e.g. @xmath129 and @xmath2 ) and could become stronger if the energy difference among major shells is reduced .
indeed , a comparison of our results with electron scattering measurements on @xmath131ni @xcite suggests that the n@xmath1lo interaction tends to overestimate the gaps between these major shells .
note that in the present calculations the @xmath130 quasiparticles are found at energies of about @xmath132 mev and overlap with the fragmented @xmath2 states .
far from the fermi energy @xmath77 , the mixing with complex configurations becomes strong and it is no longer possible to identify sharp quasiparticle and quasihole states .
still , the energy region occupied by the major shells can be identified clearly .
the n@xmath1lo interaction places the states associated with the @xmath129 shell between @xmath133 and @xmath134 mev , while the @xmath135-shell states appear below @xmath136 mev .
other hole fragments are observed around @xmath134 mev for the @xmath95 , @xmath94 and @xmath96 partial waves .
these originate from particle states in the @xmath2 shell that are partially occupied due to the smearing of the fermi surface .
nucleon knockout from these orbits requires little energy transfer and leads to low - lying states in @xmath112ni or @xmath112cu .
these states originate from the mixing of particle orbits with two - hole - one - particle configurations ( @xmath137 ) in the dyson equation . still , the @xmath112ni ground state is strongly influenced by the @xmath93 hole component .
analogous fragmentation patterns extend to the shells further away from the fermi surface , although these are not shown in fig .
[ fig : spctfnct ] . on the particle side , @xmath138 mev marks the threshold for the single - particle continuum in the @xmath139-nucleon system . above this , the exact spectral function becomes a continuous function of energy . in the present calculations a structure of separate poles is found due to the discretization of the model space .
a continuum distribution also develops for the hole part of the exact spectral function below the energy @xmath140=@xmath141 , where @xmath142 is the one - nucleon separation energy from the ground state of @xmath128 particles .
the distribution on both sides of the fermi surface is similar but not fully symmetric , the strength being stronger at large positive energies .
this is because the @xmath139-nucleon system can access a larger phase space than a single hole within the @xmath56-nucleon ground state .
this asymmetry is already observed at the level of the self - energy @xmath143 , a result in line with available fits of global optical potentials @xcite . for the proton case ,
the poles of @xmath71 correspond to the addition ( removal ) of a proton to the eigenstates of @xmath4co ( @xmath112cu ) .
the corresponding spectral strength is substantially the same as that discussed for neutrons due to the almost exact isospin symmetry of the nuclear force .
however , it is shifted to higher energies by the coulomb repulsion .
the aim of this work has been to extend large - scale calculations of self - consistent green s functions to medium mass nuclei and to investigate the properties of the single - particle spectral function of @xmath0ni .
many - body green s functions hold a number of interesting mathematical properties .
since one aims at obtaining excitations relative to a reference nucleus calculations scale more gently when increasing the number of particles as opposed to direct large - scale diagonalization methods .
only connected diagrams are summed to all orders so that the extensivity condition is satisfied @xcite .
moreover , the self - consistent approach provides a path to ensure the conservation of basic macroscopic quantities .
however , the greatest advantage of the self - consistent green s function formalism is that its building blocks , the many - body propagators , contain information on the response to several particle transfer and excitation processes .
therefore , they can be directly compared to a large body of experimental data . due to these characteristics
the formalism can be used to gain unmatched insights into the many - body dynamics of quantum mechanical systems . within this framework , the faddeev random phase approximation method proposed in ref .
@xcite is a good candidate to pursue _ ab initio _ studies of medium mass isotopes . in this work
we have presented the basic details of calculating the faddeev random phase approximation expansion and discussed results for the spectral function of @xmath0ni .
this is the first application of the self - consistent green s function approach to the @xmath2 shell region .
the calculations employ the chiral n@xmath1lo two - nucleon interaction , with a modified monopole to account for missing many - nucleon forces .
in addition to this one - parameter modification of the hamiltonian , the only remaining parameters that enter our calculations are those defining the nucleon - nucleon interaction .
calculations were performed in models spaces including up to ten major oscillator shells .
these large spaces are large enough to allow for a sophisticated treatment of long - range correlations .
the quasiparticle and quasihole energies of the @xmath2 valence orbits were found to be rather well converged . in the largest calculations they appeared to be almost constant for oscillator frequencies in the range @xmath144 $ ] mev .
these convergence properties are possible thanks to a prediagonalization of the effects of short - range correlations .
this is done using the @xmath46-matrix technique @xcite to resum ladder diagrams outside the model space .
our results put in evidence the strong sensitivity of spectroscopic factors on the particle - hole gap at the fermi surface . for the @xmath6 and @xmath7 subshell closures the bare n@xmath1lo
potential fails in describing the experimental gap ( in an analogous way to other realistic two - nucleon interactions @xcite ) .
this effect has been attributed to missing three - nucleon interactions @xcite .
it is found that a proper correction of few monopole terms of the hamiltonian allows us to extract reliable results for the fragmentation of single - particle strength .
fully self - consistent faddeev random phase approximation calculations have till now only been presented for @xmath17o .
the extension to accurate _ ab initio _ calculations in the @xmath2 shell represent a major technical advance .
however , no substantial use of parallel computation has been made in applying this formalism .
improvements in numerical algorithms are still possible and it is expected that they will allow a better treatment of fragmentation in the self - consistent approach , as well as pushing the limits of present calculations well beyond mass @xmath88 .
another obvious extension is the inclusion of explicit three - nucleon forces . within the framework of self - consistent green s function theory
this has already been achieved for nuclear matter studies @xcite .
similar developments can be expected for finite nuclei as well . for open shell systems with weakly bound states and/or resonances
, one needs a single - particle basis which can handle continuum states , as done in @xcite .
normally , this leads to a much larger space and may require parallelization of our codes .
on the other hand , the effective interaction among valence - space quasiparticles is already generated in the present calculations and can be used for standard shell - model calculations in one or two major shells .
the issue of degenerate unperturbed states for open shells systems has also been addressed for green s functions theory in ref .
@xcite by using a bogoliubov - type quasi - particle transformation . in a self - consistent treatment , one may improve on this approach by extending the faddeev - rpa method to include explicit configuration mixing between the nucleons inside the open shell .
the @xmath6 and @xmath7 subshell closure has also attracted recent experimental interest following the discussion of whether the low - lying states of @xmath4cu are strongly fragmented due to a soft @xmath0ni core , see for example ref .
while no direct experimental information is available for the transition between these two isotopes , the spectroscopic factor for neutron knockout from @xmath4ni has been measured in ref .
the present calculations describe well the quenching of the experimental cross section . at the same time
, we report predictions for both proton and neutron transfer to the other valence orbits around @xmath0ni . these calculations can thereby provide theoretical benchmarks for the forthcoming experiments of refs .
these spectroscopic factors are all in the range of 60%-70% of the independent - particle model value and in fair agreement with the observation of valence states in several stable nuclei @xcite .
the fragmentation pattern of valence orbits predicted by the faddeev random phase approximation is also found in substantial agreement with what is known for closed shell nuclei @xcite and supports the description of @xmath0ni as a doubly magic nucleus .
finally , we note that the effects of admixing configurations with several particle - hole excitations are not included in the present study .
these effects can be accounted for by using configuration interaction ( shell - model ) methods . however , based on the analysis of refs .
@xcite these corrections are not expected to be dominant .
j. lee , m. b. tsang , w. g. lynch , m. horoi , and s. c. su , preprint arxiv:0809.4686 ; m. b. tsang _ et al _ , preprint arxiv:0810.1925 and national superconducting cyclotron laboratory , michigan state university , proposal 06035 .
c. barbieri , c. giusti , f. d. pacati , w. h. dickhoff , phys . rev . * c70 * , 014606 ( 2004 ) .
d. g. middleton et .
al . , eur .
j. * a29 * , 261 ( 2006 ) .
c. barbieri and b. k. jennings , phys . rev . * c72 * , 014613 ( 2005 ) | the single - particle spectral function of @xmath0ni has been computed within the framework of self - consistent green s functions theory . the faddeev random phase approximation method and
the g - matrix technique are used to account for the effects of long- and short - range physics on the spectral distribution .
large scale calculations have been performed in spaces including up to ten oscillator shells .
the chiral n@xmath1lo interaction is used together with a monopole correction that accounts for eventual missing three - nucleon forces .
the single - particle energies associated with nucleon transfer to valence @xmath2 orbits are found to be almost converged with respect to both the size of the model space and the oscillator frequency .
the results support that @xmath0ni is a good doubly magic nucleus .
the absolute spectroscopic factors to the valence states on @xmath3 are also obtained . for the transition between the ground states of @xmath4ni and @xmath0ni
, the calculations nicely agree with heavy - ion knockout experiments . |
0.8 cm absorption - line features in the spectra of distant quasars arising from intervening damped ly@xmath0 systems are presently yielding extremely interesting information about the evolution of the gaseous content of the universe .
surveys carried out at optical wavelengths indicate that the damped ly@xmath0 absorbers contribute a cosmological mass density roughly comparable to the density of luminous matter in present - day spiral galaxies ( wolfe et al .
more recently , lanzetta et al .
( 1993 ) have analyzed a set of international ultraviolet explorer ( iue ) spectra in order to extend the search for damped ly@xmath0 absorption systems to lower redshifts .
taken together , these surveys indicate that the cosmological mass density of neutral gas associated with damped ly@xmath0 absorption decreases significantly from @xmath5 to @xmath6 , and this evolution results from a steady decrease in the incidence of high column density absorption systems with decreasing redshift .
the damped ly@xmath0 absorbers are prime candidates for the progenitors of spiral galaxies .
detailed absorption - line modelling of individual damped ly@xmath0 absorption systems indicates a resemblance to extended hi disks ( wolfe et al .
further evidence for the disk hypothesis has also come using a variety of other observational techniques .
vlbi investigation of the @xmath7 21 cm line in the quasar pks0458 - 02 reveals that the absorber extends at least @xmath8 kpc transverse to the line of sight ( briggs et al .
in addition , wolfe et al . (
1992 ) have acquired deep ccd images of the field surrounding the qso h0836 + 113 and find a ly@xmath0 emitter corresponding to the z=2.466 damped ly@xmath0 line with linear dimensions @xmath9 kpc@xmath10 , comparable to the size of a present - day disk galaxy . if the damped ly@xmath0 systems are indeed the progenitors of galactic disks , then the observed evolution of these systems with redshift places important constraints on theoretical models of galaxy formation .
most currently popular cosmological models , including cold dark matter ( cdm ) models and variations thereof , predict that structure forms in a hierarchical fashion , with small objects merging to form larger and larger structures .
gas , which is trapped and heated within a collapsing dark matter halo , may then cool and collapse to form a rotationally supported disk at the centre of the halo .
this picture has been investigated in a number of papers using analytic techniques based on the press - schechter formalism ( white & frenk 1991 ; lacey & silk 1991 ; kauffmann et al . 1993 ; lacey et al . 1993 ; cole et al .
it has also been studied using numerical techniques such as smoothed particle hydrodynamics ( sph ) ( katz & gunn 1991 ; katz , hernquist & weinberg 1992 ; navarro & white 1994 ; evrard , summers & davis 1993 ) . in this paper
, we use the most recent data of lanzetta et al .
( 1993 ) on the statistics of damped ly@xmath0 absorption systems to constrain cosmological models and the history of star formation in galaxies .
we make use of the analytic formalism developed by kauffmann & white ( 1993 ) and kauffmann et al .
( 1993 ) to keep track of the cooling of gas and formation of stars in an evolving distribution of dark matter halos in a hierarchical universe . in section 3
, we show that the rapid increase in the cosmological mass density of neutral gas at high redshifts can not be accommodated in _ any _ model where stars form with the same efficiency at all epochs . in most studies of galaxy formation in a hierarchical universe
, stars are assumed to form at a rate proportional to the amount of gas which cools at any given time .
this leads to neutral gas mass densities that remain constant or even decrease at high redshift .
we also show that a mixed dark matter ( mdm ) model with @xmath2 is not able to account for the large amounts of hi gas observed at high redshift in damped ly@xmath0 systems , simply because structure forms too late in this model . in section 4 , we construct a model in which the efficiency of star formation in the universe decreases as a power of the redshift @xmath11 , and we assume that gas cooling within a dark matter halo forms an exponential disk with a scale length proportional to the virial radius of the halo .
we find that our model is in good agreement with the available observations .
finally , we show that our empirically - derived star formation law leads to disk galaxies with spectral properties similar to those of present day spiral galaxies .
the semianalytic model of galaxy formation that we use has been described in detail in a recent paper by kauffmann , white & guiderdoni ( 1993 , hereafter kwg ) .
we will only present a brief summary of the model here .
we use an extension of the press - schechter theory due to bower ( 1991 ) and bond et al .
( 1991 ) to model the evolution of dark matter halos in a hierarchical universe .
we construct monte carlo realizations of the merging history of a present - day dark matter halo of given mass .
this allows us to trace the merging path of every halo mass element through all the progenitor halos from which the present object formed . let us now consider a set of dark matter halos present in the universe at some redshift z. for simplicity , we model our halos as truncated , singular , isothermal spheres .
we assume that when a halo forms , gas is shock - heated to the virial temperature of the halo and relaxes to a distribution that exactly parallels that of the dark matter .
gas is then able to cool and collapse to form a galaxy at the centre of the halo . in the terminology adopted by kwg
, this galaxy is referred to as the _ central _ galaxy of the halo .
the rate at which cold gas settles onto the central galaxy of a halo of circular velocity @xmath12 is given by min@xmath13 .
the infall rate @xmath14 onto the halo is given by @xmath15 where @xmath16 is the fraction of the critical density in the form of baryons and @xmath17 is the fraction of the baryons in the form of gas . in this paper , we adopt @xmath18 , which is the upper limit of values allowed by big bang nucleosynthesis if @xmath19 km s@xmath4 mpc@xmath4 ( olive et al .
the cooling rate @xmath20 can be expressed as a simple inflow equation , @xmath21 where @xmath22 is the radius at which the cooling time in the halo is equal to the age of the universe .
it is not easy to predict exactly at what rate this cold gas will turn into stars .
kwg choose a star formation law given by the simple equation , @xmath23 where @xmath24 is the total mass of cold gas , @xmath25 is a characteristic dynamical timescale for the galaxy , and @xmath0 is an adjustable parameter .
similar star formation laws have also been adopted by other authors ( white & frenk 1991 ; cole et al .
1994 ) . for a standard scalo ( 1986 ) imf ,
the number of supernovae expected per solar mass of stars formed is @xmath26 .
the kinetic energy of the ejecta from each supernova , @xmath27 , is about 10@xmath28 ergs . if a fraction @xmath29 of this energy is used to reheat cold gas to the virial temperature of the halo , the amount of cold gas lost to the interstellar medium in time @xmath30 can be estimated as , @xmath31 the parameters @xmath0 and @xmath29 together control the luminosity and cold gas content of a galaxy . in kwg ,
these parameters were fixed so that the central galaxy of a @xmath32 km s@xmath4 halo at z=0 had a luminosity and gas content equal to that of the milky way .
let us now consider this set of halos at some later redshift , @xmath33 .
our monte carlo merging histories tell us which halos present at redshift @xmath11 have merged with each other to form a larger object .
the central galaxy of this new halo is identified with the central galaxy of its largest predecessor .
the galaxies that have formed inside the smaller predecessors become _ satellites_. these satellites merge with the central galaxy at a rate governed by the dynamical friction timescale ( see kwg for more details ) .
in this section , we place constraints on cosmological models and on the star formation history of galaxies using the observed evolution of the hi mass density contributed by damped ly@xmath0 systems .
we consider three cosmological models : 1 . a b=1.5 cdm model .
this is the lowest value of @xmath34 allowed by the observed abundance of rich clusters ( white et al .
a cobe normalized model would have b=1 .
2 . a b=2.5 cdm model .
3 . a mdm model with @xmath2 normalized to cobe .
this value of the neutrino mass density has been shown to provide the best fit to the observed large scale structure and streaming motions of galaxies ( klypin et al .
1993 ; pogosyan & starobinsky 1993 ) . in all three cases ,
we assume @xmath35 and @xmath19 km s@xmath4 mpc@xmath4 . in our formalism
, the influence of the choice of cosmological model is reflected by the way the mass distribution of dark matter halos changes as a function of redshift .
structure forms progressively later as we proceed from model 1 to model 3 in the list above . in the mdm model , galaxy - sized halos only become abundant at redshifts less than 1 . the equations governing cooling and star formation within the halos
have already been described in section 2 .
note that equation ( 3 ) implies that stars always form at a rate proportional to the mass of cold gas in the halo . in figure 1
, we plot the total mass density of cold gas contained in halos with circular velocities between 50 and 300 km s@xmath4 , corresponding to the circular velocities of present - day spiral galaxies . as can be seen , the predicted @xmath1 matches the observations at low redshift in all three models .
this is not unexpected because we have tuned the parameters @xmath0 and @xmath29 in equations 3 and 4 to match the gas mass of a milky way - type galaxy at the present day .
however , @xmath1 does not increase with redshift in any of these models as in the data .
it either remains flat or , in the case of the mdm model , drops off at high redshift .
it is clear than in order to reproduce the observed increase in the hi mass density , stars must form less efficiently at high redshift .
we now show that structure formation occurs too late in the mdm model to account for the large mass of high column - density neutral hydrogen observed at large redshifts , even if we neglect star formation completely .
the short - dashed / dotted line in figure 1 shows the mass density contributed by _ all _ baryons in halos with circular velocities between 50 and 300 km s@xmath4 .
recall that we have chosen @xmath18 .
this curve falls below the observational data at redshifts greater than 2 .
the hypothesis that damped ly@xmath0 systems are the progenitors of spiral galaxies is simply not viable for an mdm cosmology .
we can make an even stronger case against mdm by considering the mass density contributed by baryons in halos with @xmath3 km s@xmath4 ( the long - dashed / dotted line in figure 1 ) .
this curve falls below the observations at @xmath36 .
rees ( 1986 ) has shown that an external uv background of an intensity 10@xmath37 erg @xmath38 s@xmath4 hz@xmath4 sr@xmath4 , sufficient to satisfy the gunn - peterson constraints on the ionization of the diffuse intergalactic medium and explain the proximity effect at redshifts @xmath39 ( lu et al .
1991 ) , would cause gas to be stably confined in halos with @xmath40 km s@xmath4 .
gas would then be unable to dissipate and collapse to form a system with high column - density .
mdm models thus appear to be ruled out even if damped ly@xmath0 systems have no connection with present - day disk galaxies .
this is not true of either of the two cdm models we have considered .
in this section , we present a model which can fit the observations .
we modify the star formation law in equation 3 to allow for a star formation efficiency which decreases with redshift , and we focus on the b=1.5 cdm model .
we now write @xmath41 where @xmath42 is the redshift at which star formation `` switches on '' .
the heavy solid line in figure 1 shows the evolution of @xmath1 for a b=1.5 cdm model and the parameters @xmath43 , @xmath44 and @xmath45 .
we are now able to reproduce the observed increase in @xmath1 at high redshift .
these parameters also leads to a `` milky way '' gas mass and luminosity that matches observations .
next we investigate whether we can reproduce the observed distribution of damped ly@xmath0 absorber column densities .
we assume for simplicity that gas cooling in a halo collapses to form a planar disk with a radial hi profile @xmath46 when viewed face on .
we also assume that this profile has exponential form : @xmath47 , where @xmath48 is the central column density of the disk .
bosma ( 1981 ) has made a detailed study of the radial distribution of hi gas in spiral galaxies .
he finds that n(hi ) typically decreases from a few times @xmath49 @xmath38 at the centre of the disk to @xmath50 @xmath38 at about 1.5 holmberg radii .
the holmberg radius for a typical @xmath51 galaxy is about 15 - 18 kpc .
this implies an hi disk scale length of about 8 kpc . we have adopted a simple scaling relation for the scale length of disks forming inside a halo of circular velocity @xmath12 at redshift @xmath11 : @xmath52 this is motivated by the relationship between the virial radius of a dark matter halo and its circular velocity ( white & frenk 1991 ) @xmath53 following fall & pei ( 1993 ) , the column density distribution at redshift z may be written @xmath54 dv_c,\ ] ] where @xmath55 is the number density of halos with circular velocity in the range @xmath56 and @xmath57 for @xmath58 and @xmath59 for @xmath60 .
the resulting column density distributions for 3 different redshift bins are shown in figure 2 .
we see that our exponential disk model is able to provide a reasonable match to the observations , including the increase in the number of high column density systems with redshift . in figure 3a
, we show the star formation rate as a function of redshift for a milky way - type galaxy .
the star formation rate increases rapidly , peaking at 8 @xmath61 yr@xmath4 at a redshift 1 and then declining to about 2 @xmath61 yr@xmath4 by the present day as all the cold gas is used up in the galaxy .
we have used the spectral synthesis models of bruzual & charlot ( 1993 ) and a scalo imf to compute the spectrum of a galaxy with this star formation history at the present day . in figure 3b
we compare the resulting spectrum with the observed spectra of typical sb and sc galaxies .
the spectra are taken from kennicutt ( 1992 ) .
the model spectrum has a 4000 break and optical colours intermediate between those of the two galaxies .
we therefore conclude that our empirically - motivated star formation law ( equation 6 ) can produce disks with spectral energy distributions consistent with those of present - day galaxies like the milky way .
in this paper , we have assumed that damped ly@xmath0 systems are the progenitors of present - day spiral galaxies and we have used their observed properties to place constraints on cosmological models and the history of star formation in disks .
we have shown that the observed evolution of @xmath1 contributed by these systems implies that stars must have formed considerably less efficiently in the past .
mdm models are unable to reproduce the mass of cold gas seen at high redshift , even in the absence of star formation .
this model predicts that at redshifts greater than 3 , most of the mass in the universe is in halos too small to permit photo - ionized gas to dissipate and collapse to form high column - density systems .
we present an empirically - motivated star formation law that can reproduce most of the observed properties of damped ly@xmath0 systems in a b=1.5 cdm cosmology . using a simple model for an absorber as a planar exponential disk , we are able to match the evolution of @xmath1 and the distribution of column densities as a function of redshift . for b=1.5 cdm
, we require that the star formation efficiency decreases in proportion to the square of the redshift and that no stars form in disks at redshifts greater than 4 .
this upper limit to the redshift of star formation would be relaxed in a model where structure forms at an earlier epoch , for example in low - density models with @xmath62 or in models with higher normalization ( lower @xmath34 ) .
it is not clear what physical mechanism would cause star formation to be less efficient at high redshift .
perhaps certain processes such as satellite accretion occur more frequently in the past and are able to thicken the disk and stabilize it against perturbations .
these are issues which need to be explored more deeply if the connection between galaxy formation and the statistics of damped ly@xmath0 absorption systems is to be understood .
* figure 1 : * the evolution of the mass density of hi gas as a function of redshift .
the points with error bars are from lanzetta et al .
the long- dashed , dotted and short - dashed lines show results for a b=1.5 cdm , b=2.5 cdm and mdm model respectively , with a star formation rate proportional to the cooling rate of gas in halos . the thick solid line shows results for a b=1.5 cdm model in which stars form less efficiently at high redshift .
the short - dashed / dotted line shows the contribution from all baryons present in halos @xmath63 km s@xmath4 in the mdm model , and the long - dashed / dotted line shows the same thing for halos with @xmath3 km s @xmath4 .
* figure 2 : * the column density distribution of damped ly@xmath0 systems in three redshift bins .
symbols with error bars are from the data of lanzetta et al .
( 1993 ) and lines show the predictions of the model described in section 4 . *
figure 3 : * ( _ a _ ) the star formation rate as a function of redshift for the central galaxy of a dark matter halo with @xmath64 km s@xmath4 .
( _ b _ ) comparison of the predicted spectrum at @xmath65 ( _ thick lines _ ) with the observed spectra of two typical nearby sb and sc galaxies ( _ thin lines _ ) .
the model spectrum was computed using the bruzual & charlot ( 1993 ) spectral synthesis code and has a scalo imf with lower and upper cutoffs of @xmath66 and @xmath67 , respectively . | 0.8 cm there is accumulating observational evidence suggesting that damped ly@xmath0 absorption systems systems are the progenitors of present - day spiral galaxies .
we use the observed properties of these systems to place constraints on the history of star formation in galactic disks , and on cosmological theories of structure formation in the universe .
we show that the observed increase in @xmath1 contributed by damped ly@xmath0 systems at high redshift implies that star formation must have been considerably less efficient in the past .
we construct a model in which gas is converted into stars with an efficiency that increases with time , and we show that this model can reproduce most of the observed properties of damped ly@xmath0 systems , including the observed distribution of column densities .
we also show that the data can constrain cosmological models in which structure forms at late epochs . a mixed dark matter
( mdm ) model with @xmath2 is unable to reproduce the mass densities of cold gas seen at high redshift , even in the absence of _ any _ star formation .
we show that at redshifts greater than 3 , this model predicts that the total baryonic mass contained in dark matter halos with circular velocities @xmath3 km s@xmath4 is less than the observed mass of hi in damped systems . at these redshifts ,
the photo - ionizing background would prevent gas from dissipating and collapsing to form high column density systems in halos smaller than 35 km s@xmath4 .
mdm models are thus ruled out by the observations . -0.55 cm -0.55 cm 0.1 mm 16.75 cm 21.3 cm constraints on models of galaxy formation from the evolution of damped ly@xmath0 absorption systems + guinevere kauffmann and stphane charlot + + galaxies : formation , galaxies : evolution , quasars : absorption lines , cosmology : theory |
spin injection across interfaces is one of the crucial ingredients for the successful implementation of novel spintronic devices @xcite . for instance
, the spin - transistor proposal @xcite relies on subjecting injected spin - polarized electrons to a controllable spin precession between the ferromagnetic source and drain .
hybrid ferromagnetic / semiconductor and dilute - magnetic - semicondutor / non - magnetic semiconductor junctions constitute basic interfaces in which to investigate spin - polarized transport . however , the efficiency of spin - injection through ideal fm / nms interfaces turns out to be disappointingly small due to the large conductivity mismatch @xcite between the metallic ferromagnet and the semiconductor .
the use of spin - selective interfaces can significantly enhance injection efficiencies @xcite .
this can be accomplished by inserting spin - dependent tunnel barriers between the fm and the nms layers @xcite .
particularly promising is spin injection from a dilute - magnetic semiconductor ( dms ) into a non - magnetic semiconductor ( nms ) .
novel dms / nms junctions ( i ) minimize the conductivity mismatch problem and ( ii ) naturally incorporate spin dependency in the transmission process @xcite .
as recently demonstrated , substantial spin injection can be achieved in these mn - based heterostructures @xcite .
more recently , a novel large magnetoresistance effect has been observed in a dms / nms / dms geometries mol .
an interesting observation of ref .
@xcite is the dependence of the magnetoresistance effect on the thickness of the dms layers : the mr effect doubles with the dms thickness .
available formulas describing gmr - type effects in magnetic / non - magnetic junctions assume 1d geometries with semi - infinite magnetic layers @xcite . in this case , the expressions for the electrochemical potentials for the spin - up and the spin - down electrons contain only decaying exponentials in the magnetic regions @xcite .
the resistances in the parallel and antiparallel configurations are calculated assuming that the device extends for a spin - flip length into the magnetic contacts .
in such an approach it is not clear how to properly take into account the voltage drop across the sample . in this work
we consider dms / nms / dms and fm / nms / fm 1d systems with _
magnetic layers , fig .
as detailed below , we describe diffusive transport in these structures via the usual diffusion theory of van son et al .
@xcite with proper physical boundary conditions between the several magnetic / non - magnetic interfaces in the system .
the formulas for the magnetoresistance obtained within our treatment are in general different from the ones previously obtained in literature , even in the experimentally relevant regime @xmath1 and @xmath2 ; @xmath3 denote the spin - flip lengths in the dms and nms layers , respectively , @xmath0 is the length of each dms layer and @xmath4 is the length of the nms layer .
however , in the case of a fm / nms / fm system where the conductivity mismatch is large , this difference is not important since the correction we find is very small in this case .
for dms / nms / dms systems , on the other hand , the conductivity mismatch is small and the magnetoresistance effect can be large .
our formulas significantly deviate from the earlier ones in this case , especially in the regime where @xmath5 , which can be magnetic - field dependent , is comparable to @xmath0 .
let us denote by 1,2 , and 3 the regions corresponding to the dms ( or fm ) , nms , dms ( or fm ) layers , respectively .
note that in fig .
1 @xmath6 corresponds to the center of the structure ( nms layer ) , @xmath7 correspond to the dms / nms interfaces and @xmath8 to the metal contact / dms interfaces .
consider first the dms / nms / dms case or , equivalently , a parallel configuration fm / nms / fm .
our task is to find the resistance of the structure @xmath9 , where @xmath10 is the total current through the structure and @xmath11 is an applied bias ( @xmath12 is the potential of the left metal contact and @xmath13 is the potential of the right contact ) . _ diffusive transport equations . _
our starting point is the diffusive transport approach of ref .
, with the basic equations relating the spin - dependent electrochemical potentials and current densities @xmath14 which hold in all three regions 1 , 2 , and 3 ( see fig .
1 ) of our system ; @xmath15 denotes the conductivity of the spin up ( or down ) electrons in the respective layer .
the difference in the spin - dependent electrochemical potential is @xcite @xmath16 where @xmath17 is the spin - flip length ( @xmath18 or @xmath19 ) in the corresponding layer . as usual , to solve the problem we first assume a general solution in each of the three regions and then match them at the many interfaces via proper boundary conditions .
below we outline this procedure , clearly stating the physical boundary conditions we use and the simplifying symmetries of the problem .
_ general solution for _ @xmath20__. _ _ consistent with eq .
( [ eq2a ] ) , we use the general expression for the electrochemical potentials in region 1 ( dms ) @xmath21 here @xmath22 is the spin flip length in the dms layer .
note the presence of the exponentially - growing term in eq .
( [ potent ] ) , since we consider a _
finite _ dms layer .
the corresponding expressions for the electrochemical potentials in the regions 2 and 3 are obtained from eq .
( [ potent ] ) by the replacements @xmath23 , @xmath24 , @xmath25 , and @xmath26 ; in addition we should make @xmath27 and @xmath28 in the nms region 2 .
the chosen form of the solution guarantees conservation of the total current everywhere in the sample . from the symmetry of the problem in the parallel configuration considered here
, it follows that the electrochemical potentials @xmath29 and @xmath30 should be odd functions of the coordinate @xmath31 , see fig .
1(b ) . in terms of the coefficients
this means that @xmath32 , and @xmath33 .
hence we can consider only the boundary conditions at the left metal contact / dms interface and left dms / nms interface .
_ boundary conditions .
_ at the metal contact / dms interface we use the continuity of the _ total _ current and the equality of the electrochemical potentials : @xmath34 .
the condition @xmath35 is consistent with assuming a metal contact with infinite conductivity .
thus we obtain the following set of equations : @xmath36@xmath37and @xmath38 at the dms / nms interface we use the continuity of the electrochemical potentials , @xmath39 , @xmath40 , the conservation of the total current @xmath10 , and conservation of the @xmath41 component across the interface
. then we have the following additional set of equations @xmath42 and @xmath43 hence we have seven equations , eqs .
( [ boundary1])([boundary6 ] ) , for seven unknown quantities @xmath44 , and @xmath45 . by solving these equations
we can determine the resistance of the structure .
this we do next .
for a fm / nms / fm and a dms / nms / dms system ( parallel configuration ) , we find for @xmath46 @xmath47 where @xmath48 , @xmath49 is the total conductivity in the nms layer and @xmath50 is the conductivity in the dms layer ; @xmath0 is the length of each dms layer and @xmath4 is the length of the nms layer . the term proportional to the magnetic - layer width @xmath0 in eq .
( [ resist ] ) is not present in earlier formulas in the literature .
note that the term proportional to @xmath51 in eq .
( [ resist ] ) differs from that in eq .
( 1 ) of ref .
@xcite because it is @xmath0 dependent and also because of its distinctive @xmath52 dependence @xcite . for the antiparallel configuration ( opposite fm layers )
, we can similarly obtain an expression for the resistance @xmath53 .
interestingly , @xmath53 can be obtained from @xmath54 in eq.([resist ] ) by changing @xmath55 to @xmath56 @xmath57 in obtaining eq .
( [ anti ] ) we have also assumed that the spin - up and the spin - down conductivities of region 3 are @xmath58 and @xmath59 , respectively . we emphasize that the term proportional to @xmath60 [ eqs .
( [ resist ] ) , ( [ anti ] ) ] does not reduce to that in ref .
@xcite for @xmath61 since the boundary conditions used here are different .
_ magnetoresistances .
_ for metallic fm / nms / fm systems we define the resistance change as usual , i.e. , @xmath62 .
the full expression is too lengthy to be shown here .
however , in the physically relevant limit @xmath63 and @xmath64 we find from eqs .
( [ resist ] ) and ( [ anti ] ) @xmath65}. \label{ourdelta}\ ] ] in deriving the above equation we have approximated @xmath54 in the denominator by its dominant contribution @xmath66 from the central nms layer .
again , our eq .
( [ ourdelta ] ) is different from the corresponding one [ eq .
( 8) ] in ref .
however , as can be seen from comparison of eq .
( [ ourdelta ] ) and eq .
( 8) of ref .
@xcite , this deviation is not really important for _ metallic _ ferromagnets .
indeed , @xmath67 which follows from our new equations practically coincides with the quantity in eq .
( 8) of ref .
@xcite for realistic values of the parameters when @xmath68 , @xmath69 , and for @xmath70 s not anomalously close to unity . in dms / nms / dms systems , on the other hand , only the parallel configuration is possible . in this case
, we define the relevant resistance change in the system with respect to the zero magnetic - field resistance @xmath71 , i.e. , @xmath72 .
we find note that @xmath71 is simply the combined nms / dms layer resistances @xmath74 , since no spin accumulation occurs for zero magnetic field . in eq .
( [ deltar0 ] ) we have indicated explicitly the additional magnetoresistance arising from the dms ( or fm ) regions through @xmath75 . note that @xmath76 goes to zero for either @xmath77 or @xmath78 .
the exact expression for @xmath79 obtained from eqs .
( [ resist ] ) and ( [ anti ] ) [ not the approximate one in eq .
( [ ourdelta ] ) ] also vanishes in these limits .
note that this vanishing of the resistance change is symmetrical in @xmath0 and @xmath52 as expected .
earlier formulas in the literature do not show this symmetry since they assume different boundary conditions .
in contrast to metallic fm / nms / fm systems , dms / nms / dms structures can have conductivities of the dms and nms regions which are comparable . besides , @xmath18 can be of the same order as @xmath0 in some experiments . experimental results in this case ,
e.g. , the _ d _ dependence of the magnetoresistance effect , should be interpreted with the use of eq.([deltar0 ] ) . as we mentioned earlier , the magnetoresistance effect observed in ref . @xcite
doubles with the dms thickness .
let us find out whether our 1d formula can explain this observed thickness dependence .
note that the _ d_-dependence in eq .
( [ deltar0 ] ) appears through the @xmath80 term ( spin accumulation effect ) .
since @xmath81 can in general be magnetic - field dependent , there is also a possible _ d_-dependence arising from the second term which describes the resistances of the dms leads . in order to investigate in more detail the _ d_-dependence in @xmath76
, we need to know the magnetic - field dependences of the quantities entering @xmath76 [ eq .
( [ deltar0 ] ) ] , particularly the magnetic field dependence of the spin - dependent conductivities @xmath82 .
unfortunately , these dependences are not known .
however , we can still gain some insight into the problem by neglecting the magnetic field - dependence of the conductivities .
in this case , the second term in eq.([deltar0 ] ) is zero and we can investigate the expected d - dependence arising from the spin - accumulation effect alone .
note that the quantity @xmath70 is null in the absence of a magnetic field and increases as a function of it .
hence , we plot @xmath76 as a function of degree of spin - polarization @xmath70 for different values of @xmath0 . figures 3(a ) , 3(b ) show plots of @xmath76 vs @xmath70 for different parameters .
we have chosen the values of @xmath83 and @xmath84 corresponding to the real experimental situation of the ref .
@xcite . from these plots
we can see that even though the absolute value of the magnetoresistance is comparable to the experimental value , the _ d_-dependence is not as pronounced as in the experiment of ref .
it is indeed true that the magnetoresistance effect is larger for increasing @xmath0 .
however , this behavior only appears at values of @xmath18 comparable to the thickness @xmath0 and even for @xmath85 @xmath86 m @xmath87 , fig .
3(b ) , the effect is way too small to explain the experimental data .
we conclude that the spin accumulation effect by itself does not explain with realistic parameters the observed _ d_-dependence of the data in ref .
@xcite .
let us mention here the possibility to explain the experiment by the magnetoresistance of the leads , see the second term in eq .
( [ deltar0 ] ) .
control experiments on a layer of the dms injector material patterned into a hall geometry showed @xmath88 negative mr @xcite .
however , the transport problem we deal with here can be shown to reduce to that of a 1d system @xcite .
note that in the case when the magnetic field dependence of @xmath89 is well pronounced and for realistic values @xmath90 and for @xmath91 the first term in eq .
( [ deltar0 ] ) is estimated as @xmath92 while the second term is of the order of @xmath93 .
thus in the case when the magnetoresistance of the leads is not zero the @xmath0-dependence associated with it can be more pronounced than the one associated with the spin accumulation effect .
_ some limiting cases .
_ from the general equations we derived above , we can obtain expressions for some interesting limiting cases . if @xmath94 , i.e. @xmath95 , we find from eq .
( [ resist ] ) @xmath96 if , in addition , @xmath97 , then @xmath98hence the resistance of the central part ( nms ) doubles as expected .
note that the resistance of the dms part does not necessarily double .
_ current spin polarization .
_ we can also easily determine some other relevant quantities in our system .
for example , the degree of spin polarization @xmath99 of the current at the dms / nms ( or fm / nms ) interface ( @xmath100 ) is equal to @xmath101 for parallel alignment .
this result coincides exactly with eq .
( 7 ) of ref .
@xcite in the limit @xmath102 . for metallic ferromagnetic contacts
we can have antiparallel alignment as well .
note that @xmath99 is an even ( odd ) function of @xmath31 for parallel ( antiparallel ) alignment . for completeness , we give below some other expressions for @xmath99 . at @xmath103 ( even function )
we have @xmath104}. \label{alpha1}\ ] ] note that for @xmath105 @xmath106 approaches the maximal possible value @xmath70 .
in addition @xmath107 thus in the middle of the sample the spin current polarization is exponentially suppressed . in the case of antiparallel configuration the corresponding values of @xmath108 at @xmath109 and @xmath110 can be obtained from formulas ( [ alpha ] ) and ( [ alpha1 ] ) , respectively , by replacing @xmath111 for @xmath112 . in conclusion , we have derived the expressions for the resistances of fm / nms / fm ( or dms / nms / dms ) structures using physical boundary conditions at the corresponding interfaces .
we have considered magnetic layers ( dms or fm ) of finite length @xmath0 . when the magnetoresistance of the contacts is negligible ( i.e. , when @xmath113 is not magnetic field dependent ) , we find that the @xmath0 dependence of the spin - accumulation contribution to the magnetoresistance is not enough to explain the experimental results of ref . @xcite .
on the other hand , in systems where @xmath113 is magnetic field dependent our formulas allow for a sizable @xmath0 dependence arising from the magnetoresistance of contacts .
interestingly , for realistic parameters @xmath114 and with @xmath91 the magnetoresistance of the contacts is larger than that from the spin accumulation effect .
s. datta and b. das , appl .
56 * , 665 ( 1990 ) .
j. c. egues , g. burkard , and d. loss , appl .
. lett . * 82 * , 2658 ( 2003 ) ; j. schliemann , j. c. egues , and d. loss , phys .
lett * 90 * , 146801 ( 2003 ) .
r. fiederling , m. keim , g. reuscher , w. ossau , g. schmidt , a. waag , and l.w .
molenkamp , nature * 402 * , 787 ( 1999 ) ; y. ohno , d.k .
young , b. beschoten , f. matsukura , h. ohno , and d.d .
awschalom , nature * 402 * , 790 ( 1999 ) .
it should be mentioned that despite the 2d geometry of the real experiment @xcite , the transport problem reduces to that of a 1d system .
this result follows from the specific current density distribution established by minimizing the joule heat ( a. khaetskii , unpublished ) . in a real experimental situation
the dms leads lie on the top of the nms layer ( at @xmath116 ) , thus the current lines are bent and the current enters the metallic contacts in the direction perpendicular to the original @xmath31-direction .
exact _ analytical treatment of this geometry , shows that the current which enters or leaves the nms layer flows only near the corners of the nms / dms interfaces .
it turns out that the relevant width @xmath117 of the dms layer where transport occurs is of the order of the nms layer height .
the current density exponentially decays for @xmath118 .
therefore we have essentially a `` bent '' 1d geometry .
note that eq.(1 ) of ref .
@xcite contains a misprint . in the numerator of this equation , instead of the first power of @xmath70
it should be @xmath60 .
the correct version of this formula can be found in : g. schmidt and l.w .
molenkamp , in _ semiconductor spintronics and quantum computing _ , ( springer - verlag , berlin heidelberg 2002 , d.d .
awschalom , d. loss , n. samarth , eds . ) .
note , however , that our eq .
( [ resist ] ) contains totally different @xmath0 and @xmath52 dependences as compared to eq .
( 1 ) of ref . | we have reconsidered the problem of spin injection across ferromagnet / non - magnetic - semiconductor ( fm / nms ) and dilute - magnetic - semiconductor / non - magnetic - semiconductor interfaces , for structures with _
finite _ magnetic layers ( fm or dms ) . by using appropriate physical boundary conditions ,
we find expressions for the resistances of these structures which are in general different from previous results in the literature . when the magnetoresistance of the contacts is negligible , we find that the spin - accumulation effect alone can not account for the @xmath0 dependence observed in recent magnetoresistance data . in a limited parameter range ,
our formulas predict a strong @xmath0 dependence arising from the magnetic contacts in systems where their magnetoresistances are sizable . |
if one could devise a mechanism that could convert even a tiny fraction of the enormous rotational energy of pulsars ( spinning neutron stars ) into particle kinetic energy , an outstanding problem of high energy astrophysics may get closer to a solution .
one could , then , provide , for instance , a possible explanation for what accelerates cosmic ray electrons to ultra high energies- a regime recently explored by the h.e.s.s .
, pamela and fermi collaborations @xcite some of the most significant observations and developments in the realm of high energy cosmic ray electrons may be summarized as : 1 ) the @xmath1 announcement of the h.e.s.s .
collaboration showing evidence for a substantial steepening in the energy spectrum of cosmic ray electrons above @xmath2 gev @xcite , 2 ) suggestions for the possible sources for the very high energy electrons ( vhe electrons ) @xcite : it is conjectured , for example , that either pulsars or the annihilation of dark matter particles might be the source of ultra relativistic energies , 3 ) the launching of the fermi spacecraft into a near - earth orbit on @xmath3 june @xmath1 - the new instruments on board detected cosmic - ray electrons up to @xmath4tev and confirmed an excess of vhe leptons in cosmic - rays @xcite , and 4 ) analysis of the observational data of the fermi telescope on millisecond pulsars ( the surface magnetic field close to @xmath5 g ) by @xcite positing that the data may have evidence of extremely high energy ( @xmath6tev ) cosmic ray electrons . this paper is an attempt to formulate a theoretical framework for particle acceleration driven by the rotational slow down of a pulsar , in particular , the crab pulsar ( @xmath7 g ) . the star rotational energy is channeled to the particles in a two step process - the excitation of langmuir waves ( via a two - stream instability ) in the electron - positron plasma in the pulsar magnetosphere ; and the damping of langmuir waves on the fast electrons accelerating them to even higher energies up to and beyond @xmath8 tev .
all known pulsars are observed with decreasing spinning rates .
the crab pulsar , for example , rotates with a period @xmath9s with a characteristic rate of change @xmath10ss@xmath11 .
the slowing down releases rotational energy that must transform into some other form .
the slowdown luminosity of the star is given by @xmath12 , where @xmath13 is the moment of inertia of the pulsar , @xmath14 cm is its radius , @xmath15 is its angular velocity , @xmath16 , and the pulsar mass @xmath17 where @xmath18 g measures the solar mass .
out of this enormous output , @xmath19erg / s , only a part @xmath20erg / s is sufficient to power the crab nebula ; the rest - almost @xmath21 of the energy budget - is released via , yet , unknown channels . in the standard pulsar model @xcite , an electrostatic field uproots particles from the pulsar s surface .
these particles are then accelerated along the magnetic field lines by the nonzero longitudinal electric field .
the electrons , accelerated to relativistic energies , begin to emit copious curvature radiation .
the emitted photons with energies beyond the pair production threshold @xmath22 ( @xmath23 is the electron s mass and @xmath24 is the speed of light ) , in turn , create electron - positron pairs thought the channel , @xmath25 ( @xmath26 is the dipolar magnetic field ) .
newly produced pairs are further accelerated and emit photons ; the cascade lasts till the nascent pair plasma , especially in regions farthest from the star , is dense enough to screen out the initial electrostatic field @xcite .
this height @xmath27 of the gap region ( where the electric field is nonzero ) is , therefore , strictly limited .
the limited accelerating region , in turn , imposes an upper limit on the maximum attainable electron energy .
the potential difference in the gap has been estimated to be @xmath28v @xcite , where @xmath29 is the dimensionless magnetic induction .
thus the potential difference , predicted by the standard gap model , may boost the electron lorentz factors to @xmath30 - not sufficient to explain the observed very high energy ( vhe ) radiation .
figure 1 shows the schematic representation of the distribution function of plasma particles in standard models of the pulsar magnetosphere .
the primary beam particles are shown by the narrow shaped area with the lorentz factor @xmath31 .
the wider region represents the distribution of secondary electron - positron pairs , characterized by a corresponding lorentz factor , @xmath32 when the distribution function has a maximum value .
several mechanisms have been proposed to , somehow , enlarge the gap and boost up energy accumulation in this zone . in particular
, @xcite proposed the so - called intermediate positronium formation mechanism , and @xcite studied the field lines supposedly curved towards the rotation and rectifying under certain conditions . invoking general relativity ,
@xcite considered the role of kerr metric in the creation of the additional electric field . in all these attempts , the gap size does increase but not enough to account for the observations .
equally noteworthy are the attempts to understand high energy acceleration in terms of the fermi acceleration mechanism @xcite .
the fermi - type acceleration is found to be efficient when the seed population of pre - accelerated electrons possesses quite high lorentz factors @xcite .
the problem of reaching such high lorentz factors , still , remains unresolved .
it is in this backdrop - searching for a mechanism for particle acceleration- that we turn to rapidly rotating pulsars .
the challenge will lie in uncovering a `` new '' mechanism for pumping out and harnessing energy from the central engine .
we explore , here , a highly efficient mechanism for transferring energy from pulsar rotation to particle kinetic energy . the invoked mechanism , first suggested ( in a limited context ) in @xcite ,
consists of two principal steps : 1 ) the time dependent centrifugal force , imparted to the relativistic plasma particles from star rotation , parametrically , drives langmuir waves , and 2 ) these waves damp on very energetic electrons to accelerate them even to higher energies . in order to appreciate the workings of the mechanism , the following short review of the standard picture of the composition of the pulsar magnetosphere may be necessary .
when acting on relativistic particles , the centrifugal force induces a periodic motion in time @xcite , and , as a result , ends up fuelling a parametric instability that may efficiently pump energy from rotation into langmuir waves .
the possibility of parametric resonance in an electron plasma was first discussed by @xcite .
it is worth noting that centrifugal effects strongly influence particle acceleration and mhd processes in the magnetospheres of pulsars @xcite and agn @xcite .
we will begin , here , by a radical and far - reaching generalization of the mechanism , first introduced in @xcite .
the refashioned model , applied to the crab pulsar , departs from the very simplified approach of @xcite that examined only a narrow subclass of centrifugally accelerated relativistic electrons- the ones with radial velocity @xmath33 @xcite .
we , instead , extend the initial particle velocity profile by letting @xmath34 with @xmath35 , the general solution allowed by the centrifugal force .
it is somewhat unexpected and rather amazing that insisting on a nonzero initial phase spawns a rather strong instability .
after this long introduction , we work out , in section 2 , the theory of centrifugally induced langmuir waves . in section 3
, we apply the model to the crab pulsar and in section 4 , we summarize our results .
to develop a model for the centrifugally induced parametric langmuir instability , we will assume that the magnetic field lines are almost rectilinear- a good approximation for distances less than the curvature radius of the field lines .
then in the rotating frame of straight field lines co - rotating with the pulsar , the metric is given by @xmath36 in the @xmath37 formalism @xcite for a single particle , the particle equation of motion , in the rotating frame of reference , reads @xmath38 where @xmath39 is the species index , @xmath40 is the charge of the corresponding particle , @xmath41 , @xmath42 is the so - called time lapse function , @xmath43 is the gravitational potential , @xmath44 is the lorentz factor , and @xmath45 and @xmath46 are , respectively , the dimensionless momentum ( @xmath47 ) and velocity . the first term on the righthand side of the equation represents the centrifugal force , and plays a crucial role in our model . to transport the equation of motion into the laboratory frame , we use the well known relation
@xmath48 connecting lorentz factors in the rotating ( @xmath49 ) and laboratory ( @xmath50 ) frames .
the identity @xmath51 takes eq .
( [ eul1 ] ) into the laboratory frame @xmath52 where we have omitted the primes for notational simplicity .
note that the first term , originating in the relativistic effects of co - rotation , has a singular behavior on the light cylinder surface ( a hypothetical zone , where the linear velocity of rotation exactly equals the speed of light ) . throughout the area of interest ,
the overall effect of the term @xmath53 will be significant .
equation ( [ eul2 ] ) is clearly fluid - like . in order to motivate this change in appearance
, we need to make a digression .
the plasma we plan to describe is really a kinetic fluid , the distribution function covers a wide range of @xmath49 ( energy ) and as will later insist , a variable phase . a proper detailed description , then , is rather involved and will require numerical modeling . in this paper , however , we plan to extract our main results in a highly reduced description .
we will assume that the plasma is multi - stream , each stream with a characteristic @xmath49 and a characteristic phase .
then we will solve the linear `` interacting '' dynamics of two such streams and demonstrate how a very interesting manifestation of the two stream instability takes place ; the phase difference between the streams will emerge as the driver . each stream , of course , is an electron - positron plasma .
equation ( [ eul2 ] ) , then , is to be viewed as equation of motion of the specie i(electron or positron ) in a given stream , and along with the continuity equation @xmath54 and the poisson equation @xmath55 completes the dynamics .
the centrifugal force @xcite is the cause of equilibrium or background velocity for the stream particles .
the electromagnetic terms do not contribute ( to the equilibrium ) because of the frozen - in condition @xmath56 . in equilibrium ,
( [ eul2 ] ) reads @xcite @xmath57.\ ] ] for highly relativistic particles ( @xmath58 ) , eq .
( [ eul0 ] ) yields the simple periodic solution , @xmath59 @xmath60 the phase @xmath61 , necessary to insure a general solution of eq .
( [ eul0 ] ) , was taken to be zero in previous work @xcite .
this was a very serious shortcoming because the zero phase solution applies only to a very small subset of particles .
the nonzero phase will insure the inclusion of all particles in the dynamics .
consequently , the instability could feed on much of the population of the magnetospheric plasma resulting in very large growth rates .
the orbits ( [ rt],[vt ] ) , along with the frozen - in condition , provide the equilibrium input for the linear stability analysis . since the calculated equilibrium is time dependent , the standard time -fourier anaylsis will not pertain- the solution of the stability problem will be more involved and nuanced . one begins with the standard splitting of the physical variables @xmath62 into equilibrium and perturbed parts @xmath63 the next step is somewhat different ; the linear perturbations are expanded as @xmath64,\ ] ] where both the amplitude and the exponent are functions of time . to keep the notation simple
, we will drop the superscript @xmath4 labelling perturbed quantities . in the linearized version of eqs .
( [ eul2]-[pois ] ) : @xmath65 @xmath66 @xmath67 @xmath68 is the electrostatic field , and i is the species index . notice that we have retained only the radial component of the linearized momentum equation , @xmath69 and @xmath70 .
the electron - positron dynamics can be combined in terms of two fluid variables : , the relative velocity @xmath71 , the differential density@xmath72 , the average fluid velocity @xmath73 , and the mean density , @xmath74 .
thus for the two stream model ( labelled by @xmath75=1,2 ) , the linear system ( [ eul3]-[pois1 ] ) takes the form , @xmath76 @xmath77 @xmath78 the two streams being coupled through the poisson equation .
we remind the reader that the two - stream model was invoked , very crudely , to capture the essence of the vast spectrum in energy and initial orbit phase available to the plasma particles .
each stream carries its own @xmath49 and phase @xmath61 .
equations ( [ eul4]-[pois2 ] ) , due to the time dependence of the orbits ( @xmath79 , and @xmath80 ) , represent a set of non - autonomous differential equations .
the goal is to determine if an arbitrary perturbation will grow in time .
we will follow two complementary methodologies to establish the existence of an instability and to determine the growth characteristics ; an exact numerical solution and an approximate analytical estimate via a quasi - modal approach .
we convert ( [ eul4]-[pois2 ] ) into two coupled equations @xmath81 @xmath82 in terms of the perturbed " densities @xmath83 where we have replaced @xmath84 by @xmath24 .
the frequencies @xmath85 are , respectively the effective relativistic plasma frequencies ( for longitudinal modes ) corresponding to the two streams , and @xmath86= b \cos ( \omega t+\phi_+ ) \ ] ] with @xmath87 , and @xmath88 in deriving eqs.([me1]-[me2 ] ) , we have invoked @xmath89 to drop the first term on the right hand side of eq .
( [ eul4 ] ) .
this parameter will turn out to be approximately the ratio of the langmuir frequency to the star rotation frequency and will lie in the range @xmath90 .
the evolution eqs.([me1]-[me2 ] ) with definitions ( [ exp1]-[exp2 ] ) have the following features : \1 ) the evolution of the two streams is coupled .
\2 ) the time dependence of the coefficients is fully contained in @xmath91 .
\3 ) @xmath91 depends on both the sum and difference of the phases associated with the two streams .
\4 ) the dependence on the phase sum @xmath92 is trivial .
in fact it can be absorbed in a redefinition of time variable ; @xmath93 will be dropped from now on .
\5 ) the phase difference @xmath94 , however , is a fundamental parameter of the two - strem system and determines the magnitude of @xmath91 .
for @xmath95 , @xmath96 , eqs.([me1]-[me2 ] ) reduce to an autonomous system with constant coefficients and lead to the modal dispersion relation @xmath97 .
the two streams collapse to one since @xmath94 , the `` measure '' of their difference , has gone to zero .
strictly speaking , this is only an approximation , since in eq .
( [ exp1 ] ) we have assumed @xmath98 , whereas @xmath99 . without this approximation
, we will get a nonzero growth rate even for @xmath100 ) .
however , this vestigial growth rate is many orders of magnitude smaller than what we will obtain for finite @xmath94 .
we would reassert that introducing the initial phase is a fundamental departure from earlier work .
it is the phase difference that differentiates the two streams and sets the stage for a strong two stream instability .
we will first perform a time fourier analysis to search for a growing quasi mode ( for non autonomous systems there are no fourier modes ) .
we must manipulate eqs.([me1 ] , [ me2 ] ) to yield @xmath101 we , then , use the bessel identity @xmath102 and take the fourier transform of eqs.([me1]-[me3 ] ) to derive the relation @xmath103 where @xmath104 equation ( [ disp1 ] ) is clearly not a dispersion relation , each fourier component ( labeled by @xmath105 ) is connected to all other components . in spite of the formidable double sum on the right hand side , we can still extract useful information that will establish the possibility of the growth of perturbations . an indicative `` dispersion relation ' ' relation
may be obtained by keeping terms with @xmath106 @xmath107 we make further analytical progress by simply keeping a single resonant term defined by the harmonic of the star rotation frequency that is in resonance with the real part of the mode frequency , @xmath108 where @xmath109 ( @xmath110 ) .
after some straightforward algebra , one can derive @xmath111 implying an imaginary part- growth rate : @xmath112 the preceding analytical calculation , though simple and revealing , should be taken only as an indicator that the langmuir waves ( charge separation perturbations ) can grow in time transferring energy from pulsar rotation to the electric field . as remarked earlier
the instability is an appropriate version of the two - stream instability in which both streams move with speeds close to @xmath24 , but whose constituent particles start with different initial orbits - the difference manifesting itself through the phase factor @xmath113 .
the expression ( [ grow1 ] ) for the growth rate tells us that , apart from its dependence on @xmath114 and @xmath115 , the growth rate is controlled by the bessel function @xmath116 of very high order @xmath117 .
bessel functions of such high order are vanishingly small unless the argument @xmath118 . for @xmath113
approaching zero , the growth rate becomes vanishingly small @xcite .
we could readily make analytical estimates for the growth rate of the rotationally driven langmuir waves for a typical pulsar .
however we must bear in mind that these estimates can be merely suggestive since the fundamental dynamics is non - autonomous in time and a given fourier mode can not capture the time behavior .
it is particularly so because @xmath119 is so large that a whole band of harmonics will contribute to the sum ( [ disp1 ] ) even if mode coupling were neglected .
however , the evolution of each mode is governed by others and the time asymptotic response does depend on the coupling of fourier modes- we must deal with the entire wave packet .
fortunately the system of eqs.([me1]-[me2 ] ) , with periodic coefficients , is readily amenable to a simple numerical solution for the densities @xmath120 , @xmath121 ( and the electric field , @xmath122 ) .
differential equations with periodic coefficients ( mathieu equation being the simplest example ) are known to display parametric instabilities in well defined regimes .
mathematica solutions demonstrate that model eqs.([me1]-[me2 ] ) are , indeed , parametrically unstable and acquire a hefty growth rate when @xmath123 ( this behavior was predicted by the analytical dispersion relation ) .
it turns out the time asymptotic behavior , for cases of interest , does settle into a pattern that growth rate of the instability can be read off from the time evolution graph .
we shall soon see that rotation ( centrifugally ) driven langmuir waves can grow at a very fast rate .
we shall now explore the detailed characteristic of the centrifugally driven langmuir waves for the crab pulsar .
our eventual aim is to put together a mechanism for delivering the rotational slow down energy to particle acceleration .
it will happen in two major steps : ( i ) in the first stage , by means of the centrifugally induced parametric instability , the pulsar s rotation energy is transferred to the electrostatic waves ; ( ii ) and in the second stage the amplified electrostatic modes undergoes landau damping on high energy electrons causing efficient particle acceleration . in order to make reasonable estimates for the instability growth rate , we recapture some details of our model . what we are studying is the instability that arises in a highly relativistic electron - positron plasma in the pulsar magnetosphere .
we have modeled the plasma in terms of two distinct streams distinguished by their @xmath49 factors and the initial orbit phase ( we have assumed the speed of both streams to be c ) .
let the two streams have quite different @xmath49s with @xmath124 .
the frequencies @xmath125 are calculated from the stream densities and @xmath126 .
assuming that energy is uniformly distributed throughout the magnetospheric plasma , we may approximate @xmath127 , where @xmath128 is the goldreich - julian number density @xcite . the magnetic field ( on the light cylinder zone )
@xmath129 is related to @xmath130 , the magnetic field close to the neutron star s surface . with these density relations ,
we conclude that one of the parameters for the numerical solution , @xmath131 will be much smaller than unity .
the most important determining parameter of the system , however , is @xmath132 , the ratio of the rotation frequency to the mode frequency ( @xmath133 ) ; the parameter @xmath134 will lie , typically , between @xmath90 , the former near the light cylinder and the latter near the pulsar surface . specifying @xmath135 , @xmath134 , @xmath136 , and the initial conditions is sufficient to seek a numerical solution .
the parameter @xmath136 is set by the wave number k and the phase difference @xmath113 .
from the general character of the differential equations , we expect strong parametric amplification when @xmath137 . from our extensive numerical study , we pick up two representative mathematica plots ( fig .
[ fig1 ] and fig .
[ fig2 ] ) . in these
, the solutions for @xmath138 , @xmath139 , @xmath140 , and @xmath141 are plotted as functions of time .
time has been normalized to the inverse of the frequency @xmath142 , the relativistic plasma frequency corresponding to the stream with lower @xmath49 and much higher density .
the respective parameters for the two figures are : for fig .
[ fig1 ] , @xmath143 , @xmath144 , and for fig .
[ fig2 ] , @xmath145 , @xmath144 .
since the @xmath146 axis denotes time measured in their respective plasma time ( inverse of @xmath142)- the `` absolute '' time scales in the two figures are different by a factor @xmath147 . for each case , we choose initial conditions(t=0 ) for what we call the sin solution : @xmath148 ; it is easy to verify that these constitute a consistent solution to the coupled differential equations in the limit @xmath149 .
the principle results of the numerical investigations may be summarized as : \a ) for initial times @xmath150(not displayed in the figures ) , the streams are essentially uncoupled and each oscillates at its own plasma frequency ; the frequencies are real .
since @xmath134 is large , this stage can be reasonably long .
for the two cases displayed , the period lasts @xmath151 plasma periods for case 1 , and @xmath152 plasma periods for case 2 . in each case
the period is @xmath153 of the star rotation period .
compared to the star rotation time , the coupling and instability build - up time is very small for the displayed examples .
b)as long as @xmath154 , even for large @xmath155 , the coupling between the streams remains small .
consequently , the perturbations amplitudes remain limited .
\c ) as @xmath136 approaches @xmath134 , the two streams begin to interact after some initial lapse of time and all 4 amplitudes begin to grow . since the two streams are assumed to have different intrinsic plasma frequencies ( @xmath156 ) the time dependence of ( @xmath157 , and ( @xmath158 are quite different .
it can be seen from the figures that although ( @xmath159 oscillate ( at a new slow frequency , the plasma oscillations are very fast and ever present - not resolved in the pictures ) and grow , the amplitudes ( @xmath158 simply grow .
interestingly enough , the slow time evolution is exponential as predicted by the floquet theory of differential equations with periodic coefficients .
thus the amplitude of entire perturbation ( not just a given fourier mode ) exponentiates , and one can define a `` global '' growth rate .
\d ) from the numerical solutions one can readily extract the the growth ( exponentiation ) time . for the solution in fig .
[ fig1 ] , it is @xmath160 plasma times @xmath161 , while the growth rate for the solution in fig .
[ fig2 ] is @xmath162 plasma times @xmath163 , where @xmath164 is the pulsar period .
thus once the instability sets in , its growth rates are very high compared to the frequency of star rotation .
\e ) for a given @xmath136 and @xmath134 , the growth rates go down as @xmath165 , precisely the scaling implied in the simple analytical estimate ( [ grow1 ] ) .
given that the growth rates are quite high , one may conclude that even the streams with quite disparate @xmath49 would produce a hefty instability .
\f ) for a moderate and not too small @xmath135 , the primary determinant of the instability gestation time , and growth rate is the relative value of @xmath166 and @xmath167 . both @xmath142 and @xmath168
are determined by the pulsar phenomenology and are independent of the wave characteristics .
thus to bring @xmath136 to the range of strong instability , one has to choose @xmath169 to be in the desirable range . in principle , then , for any finite @xmath170 , one can find a range of @xmath171 for strong growth .
the centrifugal drive for langmuir waves is , therefore , very robust and effective over a wide spectrum .
g)the two stream model calculation , though far from perfect , is perfectly capable of capturing the essence of the pumping of langmuir waves in the magnetospheric electron - positron plasma by the star rotation . a more detailed and comprehensive semi - kinetic model , that treats @xmath49 and @xmath113 as phase space variables , is certainly recommended .
however the current calculation creates a strong basis for the rapid build - up of rotationally pumped langmuir waves ( and their associated longitudinal electric fields ) so that we can confidentlly begin to explore the consequences of these electric fields for particle acceleration . after having established that the rotational energy of the pulsar can be rapidly converted into charge separation electrostatic fields
, we will now address the next issue of driving fast particles with these electric fields .
looking for a mechanism to produce super- energetic particles , we will concentrate on further energizing the most energetic particles in the pulsar atmosphere- the primary beam electrons that , to begin with , have rather high lorentz factors as high as @xmath172 @xcite . the principal energy transfer ( from waves to particles )
channel explored in this work is the damping of langmuir waves on electrons ( landau damping ) .
landau damping is , primarily , a kinetic process and will need to be appropriately integrated into what has been , till now , an essentially fluid approach . even at the cost of repetition , it helps to emphasize that , in our model , \1 ) the langmuir waves are produced by the bulk e - p plasma characterized by the relatively lower @xmath49 region of the distribution ( fig.1 ) , \2 ) while we want these waves to damp on the electron beam that appears as a small blip in the very high @xmath49 end of the spectrum ( fig.1 ) the `` producer '' and the expected `` consumer '' particles , though sharing the same physical space , are very different in their numbers , and in per particle intrinsic energy .
the study of electrostatic waves in ultrarelativistic plasmas started prior to the discovery of pulsars @xcite , but gained considerable momentum after the discovery @xcite . from this set of papers
, one learns that ( one dimensional ) electrostatic waves in relativistic e - p plasmas allow very simple dispersion in two limits : when the phase velocity is very high ( @xmath173 ) , and when the phase speed is very close to the speed of light , @xmath174 .
it is clearly the second case that is of interest to the energy transfer mechanism . in this limit ,
the real part of the dispersion relation takes the form @xcite @xmath175 where @xmath176 is a very small parameter @xmath177 denotes momentum averaging ) , and @xmath178 is the limiting value of @xmath171 for which the phase speed is exactly @xmath24 . from eq .
( [ dispr ] ) , we infer that the phase velocity will be less than the speed of light if @xmath179 ; this condition defines the regime in which such a wave could strongly interact with physical particles .
going back to the unstable langmuir waves , described in the previous section , we recall that the instability growth rates were found to be highest when the parameter ( argument of the bessel function ) @xmath180 is close to but moderately greater than @xmath119 ( the order of the bessel function ) .
the `` resonance '' condition , @xmath181 where g is number @xmath182 , implies that the phase velocity of the unstable waves , @xmath183 , is in the correct regime as long as @xmath184 is close to but less than unity . for g=1 ,
for example , only the waves generated by streams whose phase difference , @xmath185 , will be able to accelerate particles ; the highest growth pertain for waves for which @xmath186 is asymptotically close to @xmath187 and the phase velocity of such waves is asymptotically close to @xmath24 . thus the e - p bulk plasma , preferentially , generates waves with phase velocity asymptotically approaching @xmath24 .
these waves will , preferentially , react with particles that move very close to the speed of light .
in the pulsar magnetosphere , the primary beam particles ( the narrow blip on the high @xmath49 end of the distribution function shown in fig . [ fig0 ] ) , belong to the most relativistic class .
the langmuir waves , therefore , would , preferentially damp on them .
indeed , eq . ( [ dispr ] )
tells us that closer the phase velocity to @xmath24 , the smaller the value of @xmath176 , and , consequently the larger the value of required @xmath188 . to recapitulate the working of the landau damping mechanism , let us place the phase velocity of the wave in that part of the primary beam region in which the distribution function decreases with @xmath49 ( @xmath189 or @xmath190 ) .
in this zone the number of electrons that satisfy @xmath191 is more than the number of electrons with @xmath192 , @xmath193 denoting the electron speed .
less energetic particles will gain energy from the waves , whereas more energetic electrons will loose energy , but since the number of particles with @xmath191 is more than that of particles with @xmath192 , the waves will damp with a net energy transfer to the beam electrons . as an aside , the parametric instability , discussed here , may be different from the one conjured in a typical laboratory setting . in the pulsar magnetospheric situation , the plasma frequency exceeds the rotation frequency by many orders of magnitude .
the resonance is possible only for higher order harmonics and with phase velocity slightly less than the speed of light . from the relativistic plasma literature @xcite , we can read off the estimated characteristic damping rate @xcite @xmath194 where @xmath195 , @xmath49 and @xmath196 are respectively the number density , the lorentz factor and the plasma frequency of the specie on which the damping occurs .
if the instability growth rates , and the landau damping rates were very disparate , there will be little effective energy transfer from waves to the particles .
on the other hand , if the damping rates were far in excess of the growth rates , the waves will not grow much , again resulting in very little transfer from the star rotation to the waves . the most optimum scenario for an overall efficient energy pumping / transfer system
is realized when the instability growth and landau damping rates are large and comparable , @xmath197 .
since it is the electron - positron plasma that supports and feeds the two - stream instability , the corresponding characteristic energy in the unstable waves is of the order of the energy accumulated ( transferred from star rotation ) in the plasma components .
remembering that the number density associated with the e - p - plasma far exceeds that of the beam component ( see fig.1 ) , the energy gained per beam electron could be extremely large if a substantial fraction of total available " energy in plasma components were transferred via the electrostatic fields .
the condition @xmath197 is satisfied , for instance , by the choice @xmath198 , @xmath199 for the two streams .
the resulting instability growth rate is of the order of @xmath200s@xmath11 ; the corresponding growth period of @xmath201 is several orders of magnitude less than the pulsar rotation period .
the rate of energy transfer to the waves is extremely fast- the processes under consideration are extremely efficient .
let us probe further into the details of energy transfer from the star rotation into the plasma particles .
in the local frame of reference the particles are forced to slide along the magnetic field lines by means of the centrifugal force .
viewed in the laboratory frame , the reaction force @xmath202 @xcite , driving the particles , becomes infinite on the light cylinder surface .
this is a natural result- to preserve rigid rotation , the particle velocity , in this region , must exactly equal the speed of light .
note also that the radial velocity tends to zero , because on the light cylinder the particles can only rotate with the linear velocity @xmath24 .
it is clear that the maximum energy gained from the rotator can be estimated as the work done by the reaction force .
for all particles involved in the process , the energy gain may be estimated as ( @xmath203 ) @xmath204 where @xmath205 is the volume in which the pumping takes place and @xmath206 is the corresponding lengthscale .
since the aforementioned work done is transferred to the beam electrons in the same volume , the energy of a given beam particle @xmath207 can be approximately calculated by equating @xmath208 with the total energy gained by the accelerated beam particles , @xmath209 , @xmath210 one can readily show that , in the vicinity of the light cylinder surface ( @xmath211 ) , and for typical magnetospheric parameters , @xmath212@xmath213 , @xmath214@xmath213 , @xmath207 , the energy acquired by a beam electron could get to be as very high- @xmath215 of tevs or higher . because the highly relativistic particles will , inevitably , loose energy due to radiation , one must investigate how radiation will affect overall energy transfer . could radiation , for example , put a stringent limit on the maximum energy acquired ? for highly relativistic electrons and photons with @xmath216 ( @xmath217 is photon energy ) , the inverse compton mechanism operates in the klein - nishina ( kn ) regime@xcite .
energy emitted per particle per second is , then , given by the approximate expression @xmath218 @xcite where @xmath219 is the classical electron radius .
the corresponding time scale of the process is @xmath220 . using the approximate photon number density , @xmath221 ( @xmath222 is the the light cylinder radius ) , and taking into account that for the crab pulsar , @xmath223erg / s , @xmath224gev , it is straightforward to show that if @xmath225tev ( for instance ) , the aforementioned timescale exceeds the typical instability time scale ( @xmath226 ) by many orders of magnitude .
thus compton cooling is very slow compared to the wave energy transfer time .
the kn time scale goes up with the particle energy implying that at higher energies , the inverse compton mechanism in the kn channel is tot slow to impose any constraints on the maximum attainable electron energy .
the next possible loss mechanism for relativistic particles , moving in magnetic field , is the synchrotron emission with the corresponding power @xmath227 @xcite , where @xmath228 is the cyclotron frequency .
one can readily show that the electrons leaving the gap with a @xmath229 will undergo so efficient synchrotron cooling that the corresponding time scale , @xmath230 will be of the order of @xmath231s .
thus , shortly after the particles leave the gap , they radiate their transverse momentum , and very soon transit to the ground landau state . after that , zipping only along the field lines , the electrons will reach the light cylinder zone in due course of time .
it is precisely the region , where the langmuir waves , always propagating along the field lines , are excited .
therefore the wave interaction with particles will not cause pitch angle scattering , efficiently suppressing the synchrotron mechanism .
quasi linear diffusion , another possible source for imparting a pitch angle , @xcite , also does not operate because the required condition for diffusion @xmath232 , is violated for extremely energetic plasmas ( for plasmas with energy density exceeding that of magnetic field ) .
the synchrotron mechanism , therefore , is not expected to impose any constraints on the maximum attainable energies .
how about the curvature radiation emitted by particles moving along the curved magnetic field lines ?
high energy electrons with energies in ( @xmath215 of tevs ) , i.e , with particle energy density exceeding magnetic energy density by many orders of magnitude , move , practically , in a straight line .
this in turn excludes the curvature radiation from interfering with our energy transfer mechanism . as in previous cases ,
curvature radiation `` does '' not impose significant limits on @xmath207 .
the absolute upper bound on the energy that an electron may acquire , may be estimated as follows : the total power budget , the slowdown luminosity , is @xmath233erg / s , about @xmath234 of which is radiated away by the crab nebula .
if the remaining @xmath21 ) , @xmath235erg / s , were fully used up ( via the mechanism proposed in this paper ) to accelerate electrons to ultra high energies , a beam electron could emerge with an energy @xmath236 where @xmath237 is the fraction of particles involved in the process .
since the particles involved in the acceleration - radiation mechanisms , are from the region of open magnetic field lines , it is natural to assume that @xmath238 ; this bunch of particles could be boosted up to multi pev range .
this estimate should be taken for what it is- an absolute upper bound .
the most important message is that even if a very small percent of the total available energy could be transferred to fast electrons , the pulsar could be a source of ultra relativistic electrons with energies in the range of @xmath215 of tevs .
it is clear from the expression of @xmath239 that the force increases asymptotically as we approach the light cylinder zone @xmath240 .
however , the projected singularity at the light cylinder surface may be circumvented by a recently conjectured mechanism : by means of a curvature drift instability , the magnetic field lines twist and , lagging behind the rotation , lead to the force - free dynamics of particles . the rigid rotation and the subsequent violation of the causality principle @xcite is , thus , prevented .
as a result , the process will , then , stop due to the aforementioned instability .
1 . we have constructed a simple but nontrivial theoretical framework to explore the possibility of particle acceleration driven by the rotational slow down of a pulsar .
the first step in the proposed mechanism is the conversion of the rotation imparted centrifugal particle energy , via a `` two stream instability '' , to electrostatic langmuir waves in the electron - positron plasma residing in the pulsar magnetosphere .
landau damping of these centrifugally excited electrostatic waves on the high energy primary beam electrons transiting through the same region , constitutes the second step .
by carrying out a linear instability analysis , relevant to a non - autonomous system , it was shown that , for physically reasonable parameters ( pertinent for example to the crab pulsar ) , the characteristic timescale for the growth of langmuir waves is much less than the typical rotation timescales of particles .
this means that the invoked instability is extremely efficient and can rapidly convert the star slow down energy into the electric field energy 3 .
landau damping of the unstable langmuir waves on primary beam electrons that converts the electric field energy into particle energy is also shown to be equally rapid .
the combination creates a very efficient `` machine '' that generates ultra high energy ( multi tev to pevs ) cosmic ray electrons ; the star slow down luminosity being the ultimate accelerator .
the mechanism , developed in this paper , is rather general .
the magnetospheres of active galactic nuclei ( agns ) , for example , be just as suited to support centrifugally driven electrostatic waves . and these waves , could , equally efficiently , accelerate any charged leptons or hadrons .
abdo , a.a .
2009 , phys .
l , 102 , 181101 abramowitz , m. & stegun , i.a . , 1965 , _ handbook of mathematical functions , _ natl .
gpo , washington , d.c . , 1965 ) ackermann , m. et al .
2010 , phys .
d , 82 , 092004 aharonian , f. , et al .
2008 , phys .
l , 101 , 261104 arons , j. & sharleman , e.t . , 1979 , apj , 231 , 854 blandford , r.d . , 1980 ,
apj , 238 , 410 blandford , r.d . & ostriker , j.p . , 1980 ,
apj , 237 , 793 blumenthal , g. r. & gould , r. j. 1970 , rev .
phys . , 42 , 237 chang , j. et al .
, 2008 , nature , 456 , 362 deutsh , a. , 1955 , ann .
18 , 1 goldreich , p. & julian , w.h . , 1969 ,
apj , 157 , 869 kisaka , s. & kawanaka , n. , 2012 , mnras , 421 , 3543 lominadze j.g .
, machabeli g.z . &
mikhailovsky a.b . , 1979 , j. phys .
colloq . , 40 , no .
c-7 , 713 lominadze , d.g . ,
mikhailovskii , a.b . & sagdeev , r.z . , 1979 ,
jetp , 50 , 927 machabeli , g.z . , 1978 , svjpp . 4 , 914 machabeli , g. , osmanov z. & mahajan , s. , 2005 , phys .
plasmas 12 , 062901 machabeli , g.z . &
rogava , a.d . , 1994 ,
a 50 , 98 machabeli , g.z .
& tsikarishvili , e.g. , 1978 , svjpp . 4 , 920 muslimon , a.g . & tsygan , a.i . , 1992 , mnras , 255 , 61 osmanov , z. & rieger , f. , 2009 , a&a , 502 , 15 osmanov , z. , rogava , a.s . & bodo g. , 2007 , a&a , 470 , 395 osmanov , z. , shapakidze , d. & machabeli , z. 2009 , a&a , 503 , 19 rieger , f. & mannheim , k. , 2000 , a&a , 353 , 473 rogava , a. , dalakishvili , g. & osmanov , z. , 2003 , gregr , 35 , 1133 rybicki , g.b .
& lightman , a. p. , 1979
, radiative processes in astrophysics .
wiley , new york ruderman , m.a . & sutherland , p.g .
, 1975 , apj , 196 , 51 silin , v.p . , 1960 , zhetf , 38 , 1577 sturrok , p. , 1971 ,
apj , 164 , 529 tademaru , e. , 1973 , apj , 183 , 625 thorne , k. , price , r. & macdonald , d.a . , 1986 , _ black holes : the membrane paradigm _ , ( yale university press , new haven , 1986 )
tsitovich , v.n . , 1961 ,
zhetf , 40 , 1775 usov , v.v . & shabad , a. , 1985 , apss , 117 , 309 volokitin , a.s . ,
krasnoselskikh , v.v .
& machabeli , g.z . , 1987 , soviet journal of plasma physics , 11 , 310 | a new mechanism of particle acceleration to ultra high energies , driven by the rotational slow down of a pulsar ( crab pulsar , for example ) , is explored .
the rotation , through the time dependent centrifugal force , can very efficiently excite unstable langmuir waves in the e - p plasma of the star magnetosphere via a parametric process .
these waves , then , landau damp on electrons accelerating them in the process .
the net transfer of energy is optimal when the wave growth and the landau damping times are comparable and are both very short compared to the star rotation time .
we show , by detailed calculations , that these are precisely the conditions for the parameters of the crab pulsar .
this highly efficient route for energy transfer allows the electrons in the primary beam to be catapulted to multiple tev ( @xmath0 tev ) and even pev energy domain .
it is expected that the proposed mechanism may , partially , unravel the puzzle of the origin of ultra high energy cosmic ray electrons . |
the inclusive decay @xmath0 is an important example of a flavor - changing neutral current process , which has been used to test the flavor sector of the standard model .
many groups have worked on improving the theoretical analysis of this process so as to keep pace with refinements in the measurements of its branching fraction .
the effective weak hamiltonian at next - to - next - to - leading order ( nnlo ) has been obtained by calculating multi - loop matching coefficients and anomalous dimensions @xcite . while the fermionic nnlo corrections to the @xmath6 matrix elements have been known for some time @xcite , complete nnlo corrections
are presently only available for the electro - magnetic dipole operator @xcite .
however , an approximate result for the nnlo charm - penguin contributions has just been published @xcite . combining these ingredients , a first estimate of the @xmath0 branching
ratio at nnlo has been presented in @xcite .
a complication in the analysis arises from the fact that measurements of the @xmath0 branching fraction impose stringent cuts on photon energy ( defined in the @xmath5-meson rest frame ) , @xmath7 , with @xmath8 in the range between 1.8 to 2.0gev .
the standard treatment is to extrapolate different measurements to a common reference point @xmath9gev using phenomenological models @xcite . in that way ,
the experimental world average @xmath10 has been derived @xcite .
the first error is statistical , the second one systematical , the third one is due to the extrapolation from high @xmath8 to the reference value , and the last error accounts for the subtraction of @xmath11 background . a theoretical result for the branching ratio with a cut at @xmath9gev has been derived in @xcite using two - loop calculations of the photon - energy spectrum in fixed - order perturbation theory @xcite .
it has been argued that the extrapolation from the total to the partial branching fraction does not introduce additional theoretical uncertainties .
this assertion is questionable because of the dynamical relevance of a soft scale @xmath12gev , whose value is significantly lower than the @xmath13-quark mass .
accounting for the photon - energy cut properly requires to disentangle contributions associated with the hard scale @xmath14 , the soft scale @xmath15 , and an intermediate scale @xmath16 set by the typical final - state hadronic invariant mass .
when the cut value @xmath8 is chosen sufficiently low , @xmath17 becomes a short - distance scale , and renormalization - group ( rg ) improved perturbation theory can be employed to calculate the effects of the photon - energy cut using a multi - scale operator product expansion @xcite .
we have recently performed a systematic analysis of these effects at nnlo .
two - loop corrections at the soft scale were calculated in @xcite , while those at the intermediate scale were computed in @xcite . here , the analysis is completed by extracting the two - loop hard matching corrections from a comparison with fixed - order calculations of the photon spectrum . using this method
, we compute the fraction of all @xmath0 events with @xmath2gev with a perturbative precision of 5% . at this level of accuracy
several other , nonperturbative effects need to be evaluated carefully .
the event fraction receives hadronic power corrections @xmath18 governed by @xmath5-meson matrix elements of local operators .
the leading correction ( @xmath19 ) is known and turns out to be small , but terms with @xmath20 are presently unknown . recently ,
a new class of enhanced @xmath21 corrections to the @xmath0 decay rate has been identified , which involve matrix elements of nonlocal operators @xcite . a model analysis using
the vacuum insertion approximation indicates that these corrections affect the total decay rate at the level of a few percent . combining our result for the event fraction with the prediction for the total branching fraction from @xcite
, we obtain @xmath22 for @xmath9gev , where we have added in quadrature the uncertainties from higher - order perturbative effects ( @xmath23% ) , hadronic power corrections ( 5% ) , parametric dependencies ( 4% ) , and the interpolation in the charm - quark mass ( 3% ) .
two - loop perturbative corrections at the intermediate and soft scales significantly lower the branching fraction with regard to the fixed - order result given in @xcite , and they increase the theoretical uncertainty .
at leading power in @xmath21 , the @xmath0 decay rate with a cut on photon energy obeys the factorization formula @xcite @xmath24 where @xmath25 and @xmath26 . the function @xmath27 contains hard quantum corrections , the jet function @xmath28 describes the physics of the hadronic final - state jet , and the shape function @xmath29 parameterizes bound - state effects inside the @xmath5-meson @xcite . in the region of interest to our analysis , in which the quantity @xmath30gev can be treated as a perturbative scale , the double convolution integral in ( [ gammaope ] )
can be evaluated using short - distance methods @xcite . in order to isolate the effect of the photon cut , we focus on the fraction of all @xmath0 events that pass the cut @xmath31 , defined as @xmath32 . since the total decay rate can be computed in fixed - order perturbation theory at the hard scale @xmath33 , the event fraction obeys a factorization formula of the same form as ( [ gammaope ] ) , but with a different hard function @xmath34 in place of @xmath35 .
contributions associated with different mass scales can be separated from each other by evolving the various objects in the factorization formula away from a common renormalization scale @xmath36 to different `` matching scales '' , where they can be calculated reliably using fixed - order perturbation theory . in this process ,
single and double logarithms of ratios of the different scales are resummed to all orders .
the matching scales should be taken close to the default values @xmath37 for @xmath27 , @xmath38 for @xmath28 , and @xmath39 for @xmath29 .
an elegant expression describing the rg evolution of the jet function was derived in @xcite .
it involves an associated jet function @xmath40 , which is related to @xmath28 by a laplace transform .
the same technique can be applied to describe the evolution of the shape function in terms of an associated soft function @xmath41 . inserting these results into ( [ gammaope ] )
, the integrations over @xmath42 and @xmath43 can be performed , leading to @xmath44 + \delta f(e_0 ) , \end{aligned}\ ] ] where @xmath45 . despite appearance ,
this result is independent of the choice of @xmath36 and of the three matching scales @xmath33 , @xmath46 , and @xmath47 .
for the special case @xmath48 , relation ( [ master ] ) coincides with a formula derived previously in @xcite .
the terms in the first two lines arise from the rg resummation of single and double logarithms . the precise form of the evolution factor @xmath49 will be given in @xcite .
the exponent @xmath50 is an integral over the cusp anomalous dimension .
the functions @xmath34 , @xmath40 , and @xmath41 in the third line contain the matching corrections at the hard , intermediate , and soft scales , respectively .
the two - loop expression for the jet function @xmath40 has been obtained in @xcite , while the two - loop result for the soft function @xmath41 can be deduced from @xcite . in the argument of these functions , @xmath51 means a derivative with respect to the quantity @xmath52 .
the hard function can be derived by matching ( [ master ] ) with the fixed - order expression for the photon spectrum derived in @xcite . at the default scale @xmath37
, we find @xmath53 \nonumber\\ & & \mbox{}+ c_f \left ( \frac{\alpha_s}{4\pi } \right)^2 \left [ c_f h_f + c_a h_a + t_f n_f h_f \right ] , \end{aligned}\ ] ] where the term with @xmath54 is to be excluded from the sum .
the one - loop result agrees with @xcite , and the two - loop coefficients are given by @xmath55 the constants @xmath56 and @xmath57 have been obtained by numerical integration in @xcite . the complete expression for @xmath58 including scale dependence will be given in @xcite . in the last line of ( [ master ] ) , the hadronic quantity @xmath59 parameterizes the @xmath5-meson matrix element of the kinetic operator in heavy - quark effective theory .
the ellipses represent unknown hadronic power corrections of order @xmath60 .
the polynomial @xmath61 derives from the @xmath62 prefactor in ( [ gammaope ] ) .
the term @xmath63 in ( [ master ] ) collects terms that are power suppressed in the ratio @xmath64 .
these corrections are known in fixed - order perturbation theory , but no resummation is available . we express them as a perturbative series in powers of @xmath65 . the one - loop result for @xmath66 has been given in @xcite , while the dominant two - loop contribution has recently been calculated in @xcite .
for subleading contributions from other operators in the effective hamiltonian only the fermionic two - loop corrections are known @xcite ; they have a negligible impact on our results .
in ( [ master ] ) , the parameters @xmath67 and @xmath59 are defined in the on - shell scheme . to improve the perturbative behavior
, one should eliminate them in favor of appropriately defined short - distance quantities .
we use the `` shape - function scheme '' proposed in @xcite , in which low - scale subtracted heavy - quark parameters are defined via the moments of the renormalized @xmath5-meson shape function , regularized with a hard cutoff @xmath68 .
the two - loop relations between the shape - function and pole - scheme parameters have been derived in @xcite .
we adopt the conventional choice @xmath69gev and use the values @xmath70gev and @xmath71gev@xmath72 extracted from a fit to moments of inclusive @xmath5-decay spectra @xcite . [ ] [ l]@xmath73 [ ] @xmath74 [ ] [ ] @xmath75 [ ] scale dependence of the ratio @xmath76 .
the curves show the effect of varying the hard scale @xmath33 ( dotted ) , intermediate scale @xmath46 ( dashed ) , soft scale @xmath47 ( dash - dotted ) , and the reference scale @xmath36 ( solid ) by a factor of @xmath75 about their default values . in the left plot
the resummation is performed for the leading - power terms only , while in the right plot the @xmath77 term is also included ( see text for further explanation).,title="fig : " ] ref .
@xcite does not provide a result for the total @xmath0 branching fraction .
the most inclusive quantity considered is the partial branching fraction defined with a mild cut at @xmath78gev . in order to combine this prediction with our rg - improved result for the event fraction
, we define the ratio @xmath79 .
we evaluate the various matching scales at their default values , namely @xmath37 , @xmath39 , and @xmath80 , where the values of @xmath17 are different in the numerator and denominator . in order to study the residual scale dependence
, we then vary each of the four scales by a factor of @xmath81 , correlated between numerator and denominator . while the ratio @xmath76 is formally independent of these scales , the residual dependence of the truncated perturbative expression can be taken as an estimate of higher - order effects .
the results are depicted in figure [ fig : t ] .
not surprisingly , the dominant effect arises from varying the lowest scale , @xmath15 , while the scale variations at the intermediate and high scales have a lesser impact .
the term proportional to @xmath82 in ( [ master ] ) includes a subset of power corrections associated with a phase - space factor . while it was possible to perform the scale separation for these terms ,
treating them in a different way than the remaining power corrections in @xmath63 is somewhat arbitrary .
the right plot in the figure refers to the form of the factorization formula shown in ( [ master ] ) , while the left plot corresponds to expanding out @xmath83 in fixed - order perturbation theory and including it in the @xmath66 term .
while the stability with respect to variations of the soft scale is better in this case , the perturbative corrections turn out to be smaller when performing the resummation for the @xmath77 term .
the shift in central value between the two schemes is about 3% , which is inside the error bar .
this effect hints at the importance of rg resummation for the power corrections .
to quote our final result we take the average of the two schemes and assign an asymmetric error reflecting the scale variation .
this yields @xmath84 the event fraction @xmath85 receives hadronic power corrections not suppressed by inverse powers of @xmath67 , but only by powers of the soft scale @xmath17 @xcite .
these corrections are governed by @xmath5-meson matrix elements of local operators . the leading effect
proportional to @xmath59 in ( [ master ] ) is small mainly due to the smallness of its coefficient @xmath86 .
generically , we expect subleading corrections to scale like @xmath87 , for which we assign a 2% uncertainty .
the main uncertainties from parameter variations are @xmath88 for @xmath89 , @xmath90 for @xmath91 , and @xmath92 for @xmath70gev .
we also include a variation of @xmath93 due to the fact that the three - loop anomalous dimension of the shape function is yet unknown @xcite .
our value for the ratio @xmath76 is lower than the estimate @xmath94 obtained in fixed - order perturbation theory @xcite .
moreover , we find that there is a significant theoretical uncertainty inherent in the calculation of @xmath76 . in order to complete the analysis we need as input the theoretical result for the @xmath0 branching fraction with @xmath95gev , which we take from the fixed - order nnlo calculation of @xcite .
these authors find @xmath96 for @xmath78gev , where the error has been obtained by adding in quadrature the uncertainties from higher - order perturbative corrections ( 3% ) , nonperturbative effects ( 5% ) , parameter dependencies ( 3% ) , and the interpolation in the charm - quark mass employed in the nnlo estimate of charm - penguin loop graphs ( 3% ) .
the @xmath0 branching fraction receives incalculable power corrections starting at order @xmath21 , which can not be described using the operator product expansion @xcite . while we disagree with @xcite on the statement that these effects are suppressed by a power of @xmath97
, we nevertheless believe that the corresponding uncertainty should not be larger than 5% .
recently , a new class of enhanced power corrections has been identified @xcite . at tree level their effects
are parameterized in terms of @xmath5-meson matrix elements of nonlocal four - quark operators . using the vacuum insertion approximation ,
a reduction of the total branching fraction between 0.3% and 3% has been found .
accounting for this effect lowers the central value from 3.27 to 3.22 .
our final prediction for the @xmath0 branching fraction with @xmath9gev is @xmath98 where we have combined errors of the same type in quadrature .
this appears justified , since theoretical correlations in the calculations of the total branching fraction and the event fraction @xmath85 are small . in ( [ upshot ] )
we have combined all uncertainties in quadrature .
a more conservative approach would be to add the errors linearly , in which case the error becomes @xmath99 .
compared with the result @xmath100 obtained in @xcite , our central value in ( [ final ] ) is lower by about 5% and , more importantly , the perturbative uncertainty is larger by almost a factor of 2 .
both changes are a result of significant two - loop corrections encountered at the intermediate and soft scales , @xmath16 and @xmath15 .
our theoretical prediction for the @xmath0 decay rate is consistent with the present experimental world average , as is reflected in the ratio @xmath101 however , whereas for a long time the experimental result used to be lower than the theoretical one , it is now about 1.4 standard deviations larger . since in many extensions of the standard model the contributions from new physics
are expected to interfere constructively with the standard model @xmath6 amplitude , the situation has changed from one where new physics models were rather tightly constrained to one where there is now room for speculation about how the central number in ( [ npratio ] ) could be explained in terms of loop contributions containing new heavy particles .
consider , e.g. , the case of type - ii two - higgs - doublet models .
whereas in the past there used to be bounds on the charged - higgs mass from @xmath0 on the order of 500gev , the ratio ( [ npratio ] ) could now be explained with a new physics contribution from a charged higgs in just that mass range . in summary
, we have presented the first nnlo prediction for the @xmath0 branching fraction in which the effects of a photon - energy cut @xmath2gev have been properly taken into account .
low - scale perturbative corrections lower the prediction for the branching ratio and introduce a significant theoretical uncertainty even at nnlo .
our result is about 1.4 standard deviations lower than the world - average experimental value .
this re - opens the door for explorations of new physics contributions to rare flavor - changing @xmath5-decay processes .
the research of t.b . was supported by the department of energy under grant de - ac02 - 76ch03000 .
the research of m.n . was supported by the national science foundation under grant phy-0355005 .
fermilab is operated by universities research association inc .
, under contract with the u.s .
department of energy .
m. misiak and m. steinhauser , nucl .
b * 683 * ( 2004 ) 277 [ hep - ph/0401041 ] .
m. gorbahn and u. haisch , nucl .
b * 713 * ( 2005 ) 291 [ hep - ph/0411071 ] .
m. gorbahn , u. haisch and m. misiak , phys .
* 95 * ( 2005 ) 102004 [ hep - ph/0504194 ] .
m. czakon , u. haisch and m. misiak , in preparation .
k. bieri , c. greub and m. steinhauser , phys .
d * 67 * ( 2003 ) 114019 [ hep - ph/0302051 ] .
i. blokland _ et al .
_ , phys .
d * 72 * ( 2005 ) 033014 [ hep - ph/0506055 ] .
h. m. asatrian _ et al .
_ , nucl .
b * 749 * ( 2006 ) 325 [ hep - ph/0605009 ] .
m. misiak and m. steinhauser , hep - ph/0609241 .
m. misiak _ et al .
_ , hep - ph/0609232 . o. buchmller and h. flcher , phys . rev .
d * 73 * , 073008 ( 2006 ) [ hep - ph/0507253 ] .
e. barberio _
[ heavy flavor averaging group ( hfag ) ] , hep - ex/0603003 . k. melnikov and a. mitov , phys .
b * 620 * ( 2005 ) 69 [ hep - ph/0505097 ] .
h. m. asatrian _ et al .
_ , hep - ph/0607316 .
m. neubert , eur .
j. c * 40 * , 165 ( 2005 ) [ hep - ph/0408179 ] .
t. becher and m. neubert , phys
. lett .
b * 633 * , 739 ( 2006 ) [ hep - ph/0512208 ] .
t. becher and m. neubert , phys .
b * 637 * , 251 ( 2006 ) [ hep - ph/0603140 ] .
s. j. lee , m. neubert and g. paz , hep - ph/0609224 . g. p. korchemsky and g. sterman , phys .
b * 340 * , 96 ( 1994 ) [ hep - ph/9407344 ] .
m. neubert , phys . rev .
d * 49 * , 3392 ( 1994 ) [ hep - ph/9311325 ] ; phys .
d * 49 * , 4623 ( 1994 ) [ hep - ph/9312311 ] .
t. becher and m. neubert , phys .
* 97 * , 082001 ( 2006 ) [ hep - ph/0605050 ] .
m. neubert , phys .
d * 72 * , 074025 ( 2005 ) [ hep - ph/0506245 ] . t. becher and m. neubert , in preparation .
z. ligeti , m. e. luke , a. v. manohar and m. b. wise , phys .
d * 60 * , 034019 ( 1999 ) [ hep - ph/9903305 ] .
s. w. bosch , b. o. lange , m. neubert and g. paz , nucl .
b * 699 * , 335 ( 2004 ) [ hep - ph/0402094 ] .
m. neubert , phys .
b * 612 * , 13 ( 2005 ) [ hep - ph/0412241 ] .
m. b. voloshin , phys .
b * 397 * , 275 ( 1997 ) [ hep - ph/9612483 ] . | by combining a recent estimate of the total @xmath0 branching fraction at @xmath1 with a detailed analysis of the effects of a cut @xmath2gev on photon energy , a prediction for the partial @xmath0 branching fraction at next - to - next - to - leading order in renormalization - group improved perturbation theory is obtained , in which contributions from all relevant scales are properly factorized .
the result @xmath3 is about @xmath4 lower than the experimental world average .
this opens a window for significant new physics contributions in rare radiative @xmath5 decays . |
when dealing with field theory in cosmological context , one often encounters the problem that a field is not in a state which is the absolute minimum of the potential , the true vacuum , but in a metastable local minimum , the false vacuum , which is at higher free energy .
the transition from false to true vacuum proceeds by a quantum mechanical tunnelling process or by a classical transition induced by thermal fluctuations for a system at finite temperature .
the transition is local in space . during the transition a bubble is formed which is a region in space with true vacuum surrounded by false vacuum .
the bubble just large enough not to collapse is called critical bubble .
once nucleated , the bubble expands , converting false vacuum to true vacuum .
if the mass of the higgs boson is not too large the electroweak phase transition is of first order and proceeds by bubble nucleation .
this scenario is interesting because of the possibility to explain the baryon asymmetry within the minimal standard model @xcite .
a quantity of fundamental importance for a phase transition is the transition rate .
the nucleation rate per volume , @xmath0 , can be calculated from the bounce - solution of the euclidean field equation .
this solution is also called classical solution . for a theory with only one scalar field @xmath1 with a potential @xmath2 one
has to solve the equation of motion : @xmath3 the boundary condition is , that the solution @xmath4 tends to its false vacuum value if one of the @xmath5 tends to @xmath6 .
the nucleation rate can be written in the form @xcite @xmath7 where @xmath8 is the euclidean action of the classical solution @xmath9 : @xmath10 in the semiclassical approximation the factor @xmath11 is given by an expression involving functional determinants @xcite .
the negative logarithm is the one - loop contribution to the effective action .
if @xmath2 only has a single minimum the bounce - solution does not exist .
nevertheless , in some models the vacuum structure is not determined by @xmath2 alone but it is changed by radiative corrections .
these can be calculated by taking into account the quantum fluctuations of the scalar field and those fields coupled to the scalar one @xcite .
depending on the specific model , the radiative corrections yield an effective potential @xmath12 with different minima @xcite .
the bounce - solution is then calculated from a field equation that contains this effective potential
. however , the effective potential is only a part of the effective action .
it does not take into account the space - time dependence of the background fields , i. e. terms , that depend on derivatives of the background fields .
this raises the question , whether it is justified to neglect these terms in calculating the bounce @xcite . in case of the electroweak phase transition the bosonic as well as the fermionic contributions to the effective action
have been calculated , based on a bounce - solution , that was determined from the effective potential and not the effective action @xcite . in @xcite an improvement to take into account non - local effects already for the calculation of the bounce was done by modifying the kinetic term of the background field .
the radiative corrections were calculated using approximations as the derivative expansion @xcite and the heat kernel method @xcite or by exact numerical computation schemes @xcite .
the corrections due to the bosonic sector ( gauge and higgs fields ) were found to be of the same order as the classical action .
in such a case the non - local terms of the effective action are not negligible and should be taken into account when calculating the classical solution . as the fermionic corrections are small
compared to the classical action @xcite , their non - local contributions to the effective action can be omitted in the equation of motion . the standard procedure to calculate the nucleation rate works in the following way : first one adds the one - loop corrections of the effective potential to the classical action @xmath8 and gets an approximation @xmath13 of the one - loop effective action
. based upon the functional form of @xmath13 the field equation is derived and the bounce is determined .
this is used to evaluate @xmath13 and the prefactor @xmath11 of the transition rate which is given in more detail by @xmath14 here @xmath15 is the modulus of the eigenvalue of the bubble s unstable mode .
@xmath16 denotes the product of all fluctuation determinants of the bosonic sector .
aside the dominant fermionic contribution due to the top - quark which is usually taken into account for the setup of @xmath13 the fluctuation determinant due to fermions is neglected here .
the aim of this paper is to improve the calculation of the bounce which is then used as the numerical input for the evaluation of the nucleation rate .
it is no longer determined as a saddle point of @xmath13 but as a saddle point of the full one - loop effective action .
i. e. , it will be obtained as a self - consistent solution of the one - loop field equation : @xmath17 the problem is , that no analytic representation of the effective action in the one - loop approximation is known . in @xcite
a numerical method to calculate the functional derivative of an effective action has been presented .
it is based upon a computation scheme to determine full one - loop green s functions @xcite . in this paper
this technique is applied to determine a bounce - solution from a field equation where the full space - time - dependence of the non - local terms of the quantum corrections to one - loop order is taken into account .
as the numerical method to compute the functional derivative needs the background field as a numerically known input quantity the procedure we use to obtain the background field is iterative .
it finally results in a self - consistent fixed point of the iteration condition , i. e. a solution of the one - loop field equation ( [ olfeq ] ) .
a one - loop green s function is a divergent quantity that has to be renormalized .
an important issue of the presented method is the strict separation of the divergent parts , which have to be handled analytically , and the convergent part , which may be computed numerically . as no expansion in any quantity
is needed a conceptual advantage of this method in contrast to approximation schemes like the derivative or the heat kernel expansion is that the unphysical infrared divergencies these approximations suffer from do not appear . among the various models for which self - consistent solutions
are interesting this paper deals with the electroweak phase transition .
recent lattice calculations show that the perturbation theory is not reliable for higgs boson masses above @xmath18gev .
the phase transition seems to disappear at a critical higgs boson mass which is in the range between @xmath19gev and @xmath20gev @xcite and the experimental lower limit of the higgs boson mass already is beyond @xmath19gev .
hence , the generation of the baryon asymmetry at the phase transition of the standard model is ruled out already , and the interest has shifted to extensions of the standard model . nevertheless , the investigation of the first order phase transition at small higgs boson masses is still of interest and is at present actively studied on the lattice @xcite .
some aspects of the transition , like the determination of the surface tension and the latent heat , are still in a state of development .
hence , we have the possibility to develop new calculational methods in a well known model that are straightforward to extend to more complicated models .
the plan of this paper is as follows : in the next section the model will be specified and the relevant fluctuation operators are given .
the next section deals with the renormalization of the effective action that is improved by resummation . in section [ sec : numericmethods ]
the formulas for the numerical procedure how to calculate the functional derivative of an effective action and the effective action itself are derived .
finally the results are presented and discussed .
the lagrangian of the su(2)-higgs model with gauge fields and fermions in the limit of vanishing electroweak mixing angle is given by @xmath21 where @xmath22 here @xmath23 denotes the field strength tensor of the gauge fields , @xmath1 the higgs - doublet , @xmath24 the pauli matrices , @xmath25 the fermion fields of the quarks , @xmath26 the vacuum expectation value of the higgs field at temperature @xmath27 and @xmath28 the yukawa - couplings of the fermions to the higgs field .
the sum @xmath29 runs over the different flavors and colors of the quarks .
the critical electroweak bubble is a pure real higgs field configuration . on the classical level
it is @xmath30 the configuration is taken to be stationary , i. e. time independent .
as explained in the introduction the bubble profile calculated from the functional derivative of the classical action ( [ sclbounce ] ) does not exist because the tree - level potential @xmath31 only has a single minimum .
hence , instead of the classical action one has to use the effective action . for the model under consideration this effective action is given by : @xmath32=s_{\rm cl}[\phi ] + s_{\rm eff , gauge}[\phi , t ] + s_{\rm eff , ferm}[\phi ,
t]+ s_{\rm eff , higgs}[\phi , t]\;.\ ] ] this functional can not be computed exactly .
usually it is calculated in the semiclassical approximation to one - loop order . in order to calculate the one - loop approximation one needs to know the fluctuation operators .
they are found by expanding the lagrangian ( [ lstandmod ] ) around the classical field configuration ( [ classfieldconfig ] ) to second order in the small fluctuations : @xmath33 \left ( 0 \atop 1\right)\ ; , \nonumber\\ a_\mu^a(x ) & = & a_\mu^a(x)\;,\\ \psi(x)&= & \psi(x)\nonumber\;.\end{aligned}\ ] ] here @xmath34 , @xmath35 , @xmath36 and @xmath37 denote the fluctuations of the isoscalar and isovector part of the higgs field , the gauge field and the fermions respectively . the second order contribution to the lagrangian
is given by : @xmath38 one has to add a gauge - fixing term and the contributions of the faddeev - popov - ghosts to the lagrangian . here
t hooft - feynman - background gauge is used : @xmath39 from @xmath40 the fluctuation operators are read of .
they are given in euclidean metric as the bounce solution we want to obtain is a solution of an euclidean field equation .
there are three contributions of the bosonic sector to the effective action .
the first is due to the isoscalar fluctuations of the higgs field , the second is due to the coupled channel of the gauge - fields and the isovector components of the higgs field ( would - be - goldstone bosons ) and the last is due to the time component of the gauge fields and the faddeev - popov - ghosts . each of them can be written as : @xmath41 = c_i \ln { \cal j}_i = c_i \ln \det \left ( \frac{-\partial^2 + { \cal u}_i(\phi)}{-\partial^2+{\cal u}_i(0 ) } \right)\;.\ ] ] in order to shorten the notation the quantities @xmath42 are introduced and one uses that the background field is time independent .
the contribution of the top - quark is the dominant one in the fermionic sector .
therefore , nothing but its contribution is taken into account .
it is given by @xmath43 = -3 \ln \det \left ( \frac{\gamma_\mu \partial_\mu - g_t \phi } { \gamma_\mu \partial_\mu } \right)\;,\ ] ] where the factor three is due to the quark color and @xmath44 is the strength of the yukawa - coupling of the top - quark . as explained in the introduction only the leading temperature dependence of the top - quark contribution , which is part of the finite temperature effective potential , will be taken into account . hence , this calculation is not self - consistent with respect to the fermionic sector . the first step is to extract the contributions of the different fluctuating fields to the high temperature approximation of the one - loop effective potential @xcite .
based upon the fluctuation operators given above the sum of the tree level potential @xmath31 and these one - loop corrections can be written as : @xmath45\;,\nonumber\end{aligned}\ ] ] where @xmath46 , @xmath47 , @xmath48gev , @xmath49 and @xmath50 .
the form of the constants @xmath8 , @xmath51 , @xmath52 and @xmath53 differs from those in @xcite , because of the contributions of the higgs boson and would - be - goldstone boson fluctuation taken into account here .
the renormalization has been performed in the minimum of the classical potential at @xmath27 .
the given representations of the effective action are quite formal .
they are divergent and have to be renormalized .
furthermore , they have to be evaluated for a system at finite temperature .
this is done in the next section
the vacuum structure is changed due to the contributions of the different fluctuating fields to the one - loop effective potential . the tree - level mass and the tree - level self - coupling of the higgs field
are modified due to effects of finite temperature .
they will be modified again by effects of higher order .
dependent on the temperature these corrections are not negligible and a pure one - loop calculation is not reliable @xcite . in order to take into account at least the dominant contributions of the higher order effects
one has to consider the sum of one - loop - daisy - graphs , replacing the expressions due to the inserted tadpole - graphs by their leading temperature dependence which is proportional to @xmath54 .
also the vertices are replaced by those which take into account the temperature dependence of the coupling . in practice the summation of daisy - graphs is performed by replacing the tree - level - potential @xmath31 in the lagrangian by the high temperature potential @xmath55 given in the preceding section @xcite . then the fluctuation operators and the contributions to the effective action are determined .
the effective action based upon these fluctuation operators is improved by resummation .
it can be written as @xmath56=\sum_{i } s_{i,\rm r}[\phi , t]=\sum_{i } c_i \sum_{n=-\infty}^{+\infty } \ln \det \left ( \frac{-\delta + \nu_n^2 + { \cal u}_i(\phi , t)}{-\delta + \nu_n^2 + { \cal u}_i(0,t ) } \right)\;,\ ] ] where @xmath57 is given by ( [ ua0eta])ff . with @xmath58 and
@xmath59 replaced by @xmath60 expanding the effective action in a series of one - loop feynman - graphs it can be rewritten as @xmath61^{-1 } \left [ { \cal u}_i(\phi , t ) - { \cal u}_i(0,t)\right ] \right\}^k\\ & = & \!\!\!\sum_{k=1}^{+\infty } s_{i,\rm r}^{(k)}\;.\end{aligned}\ ] ] here @xmath62^{-1}$ ] is a formal representation of the green s function associated to that operator .
the superscript @xmath63 denotes the order of the corresponding graph ( see fig . 1 ) . in analogy
also the improved one - loop effective potential @xmath64 exists .
it has the following relation to the usual one - loop effective action @xmath65 : @xmath66\\ & = & c_i t \ ; { \rm tr } \ ; \sum_{n=-\infty}^{+\infty } \int \ !
\frac{{\rm d}^3p}{(2\pi)^3 } \ , \ln \left [ \mbox{\boldmath$p$\unboldmath}^2 + \nu_n^2 + { \cal u}_i(\phi)\right]\\ & & + c_i t \ ; { \rm tr } \ ; \sum_{n=-\infty}^{+\infty } \int \ !
\frac{{\rm d}^3p}{(2\pi)^3 } \ , \ln \left [ \frac{\mbox{\boldmath$p$\unboldmath}^2 + \nu_n^2 + { \cal u}_i(\phi , t ) } { \mbox{\boldmath$p$\unboldmath}^2 + \nu_n^2 + { \cal u}_i(\phi)}\right]\nonumber\\ & = & v_i(\phi , t ) + v_{i,\rm ring}(\phi , t)\;.\end{aligned}\ ] ] the divergencies in @xmath67 are temperature independent .
they are cancelled by the familiar @xmath27counter terms
. however , also @xmath68 contains divergencies which furthermore are temperature dependent .
hence , the substitution of @xmath31 by @xmath69 accounts for finite and infinite contributions of higher orders .
the latter are removed by the use of thermal counter terms @xcite .
they are not fixed by renormalization conditions but are chosen in such a way that they cancel the two divergent orders of @xmath70 exactly .
these two orders ( @xmath71 and @xmath72 ) can be read of from : @xmath73^k - \left [ \frac{{\cal u}_i(\phi ) - { \cal u}_i(0,t)}{\mbox{\boldmath$p$\unboldmath}^2 + \nu_n^2 + { \cal u}_i(0,t)}\right]^k\right\}\nonumber\end{aligned}\ ] ] however , the high temperature contribution of @xmath74 , i. e. the contribution of the static matsubara mode , is finite and is not subtracted . for the numerical application the resummed one - loop effective action is separated into a finite part that is evaluated numerically and an infinite contribution that has to be renormalized .
the latter is given by the first two orders of the series ( [ sirphitexpansion ] ) .
in performing the renormalization by adding the counter terms one finds contributions already taken into account in the high temperature potential @xmath75 ; we denote these terms as @xmath76 .
these terms must be subtracted in order to avoid double counting ; we call the resulting difference @xmath77 : @xmath78 here we make use of that the high temperature approximation of the second order contribution , @xmath79 , is finite .
therefore , it is treated together with the finite part , @xmath80 , and is evaluated numerically : @xmath81 here a superscript @xmath82 denotes the sum of contributions beginning with the @xmath83-th order . the index ` ns ' denotes the contributions of the non - static matsubara modes .
it is known to be approximated sufficiently by a gradient expansion @xcite .
it is found to be negligible small compared to the contribution due to the static mode .
therefore it is neglected and only @xmath84 is evaluated numerically .
this is postponed to the next section . using @xmath85 and neglecting the terms with @xmath86 and @xmath87 , which are small at the temperature we are working at
, the analytic part @xmath88 is found to be : @xmath89\phi^2 \right.\nonumber\\ & & + \frac{g^3 t}{32\pi } \phi^3 + \left ( \frac{g^2}{4 } + \lambda\right ) \left [ -\frac{3 \check m_h t}{8\pi } - \frac{3\check m_h^2}{32\pi^2 } + \frac{3m_w^2}{32\pi^2 } + \frac{3m_h^2}{64\pi^2 } \ln \frac{m_w^2}{a_bt^2}\right ] \phi^2\nonumber\\ & & \left .
+ \frac{g^3t}{16\pi}\phi^3 -\frac{3g^2}{32\pi^2 } \left ( \mbox{\boldmath$\nabla$\unboldmath } \phi\right)^2 \ln \frac{m_w^2}{a_bt^2 } \right \}\;.\end{aligned}\ ] ]
in the previous section the calculation of the finite contribution to the resummed effective action due to the different fluctuating fields was postponed . combining them to @xmath90 , its functional derivative with respect to the background field @xmath9
can be represented as @xmath91\nonumber\\ & = & { \rm tr } \left [ c \ ;
\frac{\delta { \cal m}^2(\phi , t)}{\delta\phi ( \mbox{\boldmath$x$\unboldmath})}g(\mbox{\boldmath$x$\unboldmath } , \mbox{\boldmath$x$\unboldmath } ) \right ] -g \ ; { \rm tr } \left [ c \ ; \mbox{\boldmath$\zeta$\unboldmath } \mbox{\boldmath$\nabla$\unboldmath } g(\mbox{\boldmath$x$\unboldmath},\mbox{\boldmath$x$\unboldmath } ) \right]\;.\end{aligned}\ ] ] where @xmath92 , @xmath93 and @xmath94 are the @xmath95 generalisations of the matrices defined in ( [ ua0eta])ff . at finite temperature .
they are given explicitly in appendix [ app : partialwave ] .
@xmath96 contains the factors of degeneracy for the different fluctuations .
its non - zero components are given by @xmath97 the green s function @xmath98 is the solution of @xmath99 g(\mbox{\boldmath$x$\unboldmath},\mbox{\boldmath$x^\prime$\unboldmath})= \mbox{\boldmath$1$\unboldmath } \cdot \delta^{(3)}(\mbox{\boldmath$x$\unboldmath } -\mbox{\boldmath$x^\prime$\unboldmath})\;.\ ] ] a method for the numerical computation of such a one - loop green s function was derived in @xcite . in order to determine
@xmath100 it is decomposed into its partial wave contributions .
the details of this calculation are given in appendix [ app : partialwave ] .
it turns out that by the decomposition one channel of the coupled system can be combined with the faddeev - popov channel so that the system reduces to a @xmath101system . extracting the order @xmath102 from ( [ funcder_genform ] ) it takes the form : @xmath103 here the radial green s functions
@xmath104 are defined by : @xmath105 the non - zero components of the potential @xmath106 are @xmath107 this coincides with @xcite where it was derived by reduction of the sphaleron system .
the radial green s function is given by @xcite @xmath108 \;,\ ] ] where @xmath109 and @xmath110 are the linear independent solutions of the homogeneous equation @xmath111 f_{k , l}^{\alpha \pm}(r)=0\;.\ ] ] the mode functions @xmath109 ( @xmath110 ) are regular ( singular ) at infinity . in order to calculate the mode functions
numerically their asymptotic behaviour at infinity is separated from the rest by the following ansatz @xcite : @xmath112 the functions @xmath113 satisfy the equation @xmath114 one obtains @xmath115 in the massive channels we obtain the modified bessel functions , while in the massless ones the solutions reduce to powers of @xmath116 .
the fifth channel only contributes in the s wave .
therefore , its bessel function is given only for @xmath117 .
the normalization constants and @xmath118 are chosen in such a way that the wronskians of the different channels become equal : @xmath119=\frac{1}{z^2}\,.\ ] ] the differential equation of the rest mode functions @xmath120 reduces to : @xmath121 h_{n , l}^{\alpha\pm}(r)=v_{nk } \left ( \delta_k^\alpha+h_k^{\alpha\pm}(r ) \right ) \frac{b_{l_k}^\pm(z)}{b_{l_n}^\pm(z)}\;.\ ] ] this equation can be used for generating the functions @xmath122 order by order in the potential @xmath123 . introducing the contribution of order @xmath83 in the potential as @xmath124 and defining @xmath125 as in @xcite ,
the order @xmath126 of the green s function is given by : @xmath127\end{aligned}\ ] ] in order to calculate the nucleation rate @xmath0 the effective action in the one - loop approximation is needed .
a method to calculate its functional derivative with respect to the background field has been given in the previous subsection .
for the purpose of renormalization the divergent parts have been separated from the rest and have been handled analytically .
the same separation can be used here in order to evaluate the effective action itself .
the analytic part is given by ( [ shrana])ff . the numerical contributions @xmath128 are determined in the following .
for the numerical computation of the effective action the fluctuation determinant @xmath129 is decomposed into partial waves @xmath130 here @xmath131 and @xmath132 are the fluctuation operators defined in ( [ gkmdefequ ] ) .
we need to know two @xmath133 matrices @xmath134 and @xmath135 which contain the @xmath136 linear independent solutions @xmath137 and @xmath138 of the system of coupled differential equations @xmath139 the boundary conditions of these functions are chosen to be regular at @xmath140 .
the lower index denotes the @xmath136 components while the different solutions are labelled by the greek upper index .
these solutions are normalized such that @xmath141 on these conditions the statement of the theorem is @xcite : @xmath142 the theorem has been applied for computing the one - loop effective action of a single scalar field on a bubble background @xcite , and of a fermion system on a similar background @xcite previously .
furthermore it has been used to calculate the bosonic fluctuation determinant of the critical bubble @xcite .
it was found to yield very precise results , in addition to providing a very fast computational method . if the theorem is applied at @xmath143 the fluctuation determinant @xmath144 is obtained .
the consideration of finite values of @xmath145 is necessary in the discussion of zero modes . for simplicity
the following formulas are given only for @xmath143 .
in analogy to the mode functions @xmath146 also the functions @xmath147 can be decomposed as : @xmath148 with @xmath149 given by ( [ eq : besselfunc ] ) .
the consequence is that the separated functions @xmath150 satisfy the same type of differential equation as the rest mode functions @xmath151 .
only the boundary conditions differ , namely @xmath152 and @xmath153 .
the two functions are related by : @xmath154 \left [ \mbox{\boldmath$1$\unboldmath } + \widetilde h_l(\infty ) \right]^{-1 } - \mbox{\boldmath$1$\unboldmath}\;.\ ] ] after determining @xmath147 one obtains the desired order of the fluctuation determinant .
@xmath155 - { \rm tr } \ , \widetilde h_l^{(1)}\right\}\;. \label{j2bar}\ ] ] as the determinant @xmath129 is a product of three individual determinants due to the coupled and uncoupled channels one has to take into account the different factors of degeneracy for each of them ( see ( [ funcder_genform ] ) ) . in order to solve the one - loop field equation iteratively we have to choose a suitable separation of the equation into a left - hand and right - hand side . then setting the right - hand side to zero results in a first approximate solution of the equation
this is taken as the start profile for the iteration .
in particular we choose the following form of the field equation @xmath156 with @xmath157^\frac{3}{2 } + \frac{3\check m_h t}{8\pi } \left [ \left ( \frac{g^2}{4}+\lambda_t \right)\phi^2 - 3et\phi \right]\nonumber\\ & & + \frac{\check m_h^3 t}{4\pi } + \frac{g^3t}{32\pi}\phi^3 + \left [ -\frac{3\lambda\check m_h t}{8\pi } + \frac{3\lambda}{32\pi^2 } \left ( \frac{m_h^2}{2 } \left ( \ln \frac{m_h^2}{a_bt^2}+2 \right)-\check m_h^2 \right ) \right]\phi^2\nonumber\\ & & + \left ( \frac{g^2}{4}+\lambda \right ) \left [ -\frac{3\check m_h t}{8\pi } - \frac{3(\check m_h^2 - m_w^2)}{32\pi^2 } + \frac{3m_h^2}{64\pi^2 } \ln \frac{m_w^2}{a_bt^2 } \right]\phi^2\end{aligned}\ ] ] i. e. , in addition to @xmath55 the contributions of the isoscalar and isovector part of the higgs field to the resummed effective potential are shifted to the left - hand side . for a suitable chosen temperature the potential @xmath158 has a secondary minimum at a nonzero value of @xmath9 .
we denote this minimum as @xmath159 .
it is used to scale the dimensional quantities as follows : @xmath160 all masses are scaled with @xmath161 .
the inhomogeneity @xmath162 consists of ( [ shrana ] ) and ( [ funcder_result ] ) reduced by those terms that were shifted to the left - hand side .
@xmath163\nonumber\\ & & -3t\left ( \lambda_t\phi -et\right ) \sum_{l=0}^{+\infty } \frac{2l+1}{4\pi } g_{h , l}^{\overline{(1 ) } } + \frac{3g^2t}{16\pi}\phi g_{a_0\eta,0}^{\overline{(1)}}\\ & & -\frac{3\lambda_t t \phi}{4\pi } \left [ \left ( 3\lambda_t\phi^2 + \check m_h^2 \right)^\frac{1}{2}-\check m_h \right]-\frac{3g^3t}{16\pi}\phi^2 - \frac{3g^2}{32\pi^2 } \delta \phi \ln \frac{m_w^2}{a_bt^2}\nonumber\\ & & -\frac{3t}{8\pi } \left [ 2 \left ( \frac{g^2}{4}+\lambda_t \right)\phi-3et \right ] \left [ \left ( \left ( \frac{g^2}{4}+\lambda_t \right)\phi^2 - 3et\phi+\check m_h^2 \right)^\frac{1}{2}-\check m_h \right]\nonumber\;.\end{aligned}\ ] ] for the purpose of comparison the solution @xmath164 of the equation @xmath165 is also calculated .
it is denoted as the classical profile . in order to avoid large changes between two subsequent profiles especially at the beginning of the iteration
we use a relaxation , i. e. the field equation ( [ fieldequation3 ] ) is written as @xmath166 by a suitable chosen @xmath167 the convergence of the iterations of the profile can be improved . for the isoscalar fluctuations there are two contributions to the effective action and its functional derivative that have to be treated individually : namely the instable mode in the partial wave @xmath117 and the translation zero mode in the partial wave @xmath168 .
a bound state occurs in the s - wave , i. e. a state with a negative eigenvalue .
as this is the only negative eigenvalue the total determinant is also negative and the one - loop contribution to the effective action is complex .
this corresponds to the fact that the state under consideration , the critical bubble , is unstable . in the p - wave
there exists a mode with the eigenvalue zero .
it corresponds to the fact that the critical bubble breaks translational invariance . because of this mode the determinant itself is zero and the effective action diverges .
for the calculation of the nucleation rate instead of the fluctuation determinant with these two properties we need one where the negative eigenvalue has been replaced by its modulus and where the three eigenvalues zero have been removed .
these changes to the determinant are indicated by two primes : @xmath169}{\det \left [ -\delta + \nu_n^2 + m_h^2(0,t ) \right]}\;.\ ] ] as three eigenvalues zero have been removed from the determinant of the isoscalar fluctuations the rate gets its dimension @xmath170 , i. e. @xmath171 .
the further dimension @xmath172 is due to the eigenvalue @xmath15 of the unstable mode .
the numerical method presented in the previous sections does not explicitly make use of the eigenvalues .
therefore , we have to show how these two contributions are removed from the effective action this has also been discussed in @xcite and its functional derivative . in case of the unstable mode
this is performed easily .
one computes the particular partial wave contributions according to ( [ j2bar ] ) . in the partial wave
@xmath117 only the sign of the determinant has to be inverted before taking the logarithm . for the functional derivative
the sign is irrelevant because due to @xmath173 the sign of each eigenvalue enters the equation twice .
the removal of the translational zero mode is more complicated .
an alternative representation for the green s function @xmath174 compared to ( [ g_represent ] ) can be deduced from the eigenvalues and eigenfunctions of @xmath175 \psi_\alpha(r)=\omega_\alpha^2 \psi_\alpha(r)\;.\ ] ] it is given by : @xmath176 obviously this green s function is divergent when the spectrum of the operator contains the eigenvalue zero . defining a new green s function as @xmath177
the divergence manifests itself in the limit @xmath178 in the form of a @xmath179pole .
as the zero mode is the only mode which contributes to this pole its contribution can be removed by defining the green s function at @xmath143 without the zero mode in the following way : @xmath180\;.\ ] ] in the special case under consideration , where the green s function only has to be calculated at @xmath181 , the practical way to subtract the zero mode in the p wave is to compute the function @xmath182 for sufficient many different values of @xmath145 .
then in the function @xmath183 the parameters @xmath11 , @xmath8 and @xmath184 are adjusted by a fit with respect to @xmath145 for each value of @xmath185 . the subtraction of the zero mode is then done by setting @xmath186 . the computation of the function @xmath182 is done in the same way as described in section [ sec : numericmethods ] .
the only difference is , in the definition of the radial part of the green s function ( [ gkmdefequ ] ) and in all consecutive equations a @xmath187 has to be added to the partial wave operator @xmath132 .
hence , the ` mass ' @xmath188 of the isoscalar channel changes to @xmath189 .
the one - loop contribution of the isoscalar fluctuation to the fluctuation determinant is determined according to ( [ j2bar ] ) in each partial wave . in the p wave the following expression has to be evaluated : @xmath190\;.\ ] ] the numerical investigation shows that @xmath191 and the logarithm diverges due to the zero mode in this partial wave . in order to subtract the mode the computational method to determine the fluctuation determinant
is extended in the same way as before in case of the green s function .
i. e. , the theorem mentioned above is applied at finite @xmath145 . in order to subtract the zero mode the determinant @xmath192 has to be replaced by @xcite @xmath193 in practice @xmath194
is computed for some sufficient small values of @xmath145 and is used to fit the constants @xmath11 , @xmath8 and @xmath184 in @xmath195 with these techniques the unstable mode and the translation mode are removed from the effective action as well as from its functional derivative . concerning the translation mode
a further remark is necessary .
an exact zero mode only exists when the profile @xmath9 is determined from the classical field equation ( [ classprofequ ] ) .
the one - loop equation differs from the classical one due to the non - vanishing right - hand side of the equation . as a consequence one finds a bound state in the spectrum in the p wave below the continuum threshold but
its eigenvalue only is near to zero but it is not exact zero .
nevertheless this mode is subtracted . in the coupled system of gauge fields and would - be - goldstone bosons
a zero mode exists in the s - wave . for the treatment of this mode we adpopt the prescription given in @xcite . in order to determine the profile for one step of the iteration a boundary problem has to be solved . at @xmath140
the slope of the profile has to vanish and at @xmath196 the function has to tend to zero exponentially .
the profile is obtained using a shooting method : the value of the profile at @xmath140 is varied until the profile gets the correct behavior at infinity .
we typically use 1000 steps on a scale from @xmath140 to @xmath197 where the latter typically is below 200 in the units @xmath198 we are working in .
the numerical technique used to integrate the differential equation is the nystrm method . for small values of @xmath185
the profile can be represented as @xmath199 .
hence , for a given profile the parameters can be obtained using a simple fit .
then for the rest mode functions @xmath200 which are calculated from the differential equations ( [ dglhnalpha ] ) the behavior at small @xmath185 can be evaluated analytically using a power series expansion . using this expansion
the functions have the correct starting properties for the following numerical integration .
hence , we can avoid numerical problems due to the divergent behavior of the bessel functions in this region .
the radial green s functions are composed from the rest mode functions according to ( [ calcghl ] ) and then the partial wave summation ( [ inhomogeneity ] ) is performed .
this sum is executed explicitly up to @xmath201 . at large
@xmath202 the product @xmath203 behaves as @xmath204 , no matter which channel @xmath205 one considers . using a fit
the coefficients @xmath206 , @xmath207 and @xmath96 can be obtained for each value of @xmath185 and then the value of the sum from @xmath208 to infinity can be estimated . as a cross - check
we start this estimate at @xmath209 and it is tested whether it approaches a constant when @xmath202 reaches @xmath208 . in order to check the precision various cross - checks have been implemented .
the equality @xmath210 = \delta_\alpha^\beta\ ] ] is satisfied numerically with a relative deviation of the order @xmath211 .
especially the wronskian between @xmath212 and @xmath213 satisfies : @xmath214 & = & \delta_\alpha^\beta + \widetilde h_\alpha^\beta(\infty)\\ & = & \delta_\beta^\alpha + h_\beta^{\alpha+}(0)\;. \label{wronski_ident2}\end{aligned}\ ] ] the relative deviations from these identities are of the order @xmath215 for ( [ wronski_ident1 ] ) and @xmath211 for ( [ wronski_ident2 ] ) .
the former being not as precise as the latter is due to the bad convergence of the functions @xmath216 at large @xmath185 . in the massless channels they only achieve their asymptotic values as @xmath217 . hence
, their value at @xmath196 only can be estimated from a fit .
on the other hand the constancy of the product of @xmath218 with the wronskian is satisfied excellently .
therefore we do not make use of this fitted quantity in the further process .
the numerical results are given in table 1 and in the figures 2 to 7 .
the higgs boson masses used here are 30gev and 40gev .
the mass of the top - quark is taken to be @xmath219gev and for the gauge boson mass we use a value of @xmath220gev .
the gauge coupling @xmath221 is taken as @xmath222 . the vacuum expectation value of the higgs field
is then fixed to be @xmath48gev .
the relevant temperature interval is determined from the approximation of the resummed effective potential @xmath158 .
it is found to be smaller than the interval determined from the high temperature approximation @xmath55 of the effective potential .
the values are given in table 1 .
within this interval we choose equidistant steps for the temperature .
we leave out the range of temperatures close to the roll - over temperature @xmath223 , because these correspond to small thick wall bubbles which are of minor cosmological interest . we give the results for the logarithm of the nucleation rate @xmath0 , i. e. the rate including the one - loop corrections and prefactors as defined in ( [ nuclrate ] ) .
it is evaluated using two different bubble profiles .
first we use the self - consistent profile , i. e. the solution of ( [ fieldequation3 ] ) .
the results have to be compared to that ones we obtain from using the profile @xmath164 , which is the solution of ( [ classprofequ ] ) .
the results are displayed in the figures 2 and 3 where we give the logarithm of the nucleation rate versus the temperature .
for the results based on the self - consistent profile we also present a function that fits the explicitly calculated points very well .
the function has the form @xmath224 the results for @xmath206 and @xmath207 depend on the specific parameter set .
we find @xmath225 and @xmath226 for @xmath227gev and @xmath228 and @xmath229 for @xmath230gev . at @xmath227gev
the self - consistently calculated transition rate is smaller than the rate obtained from the classical profile , while at @xmath230gev it is just the other way around .
this is due to the relation between the temperature intervals .
while in both cases the interval suitable for finding the self - consistent solution is located within the interval derived from the high temperature potential it is shifted towards higher temperatures with increasing higgs boson mass ( see table 1 ) . for comparison
we map the temperature intervals @xmath231 $ ] and @xmath232 $ ] onto the interval @xmath233 $ ] by introducing the dimensionless quantity @xmath234 .
the results of both parameter sets are given in figure 4 .
for both parameter sets the self - consistently obtained rate is increased compared to the ordinary one .
hence , the effect of using the self - consistent profile is an enhancement of bubble nucleation .
based on the results for the nucleation rate we can investigate some aspects of the cosmological electroweak phase transition . as the universe expands the temperature @xmath235 decreases .
once the critical temperature @xmath236 is reached critical bubbles can start to nucleate and grow . the corresponding time is obtained using the relation between time and temperature @xcite : @xmath237 hence
, the nucleation rate @xmath0 can be considered as a function of time as well . for a given velocity @xmath238 of the bubble wall the number density of bubbles of size @xmath185 at the time
@xmath239 is given by @xmath240 then the fraction of space that is still in the symmetric phase is obtained as @xcite : @xmath241\;.\ ] ] this quantity is used to define implicitly the time @xmath242 when the nucleation is completed by @xmath243 . as it marks the end of the phase transition it is a quantity that is easier accessible in lattice simulations than other characteristics of the transition .
the fraction of space that is already in the asymmetric phase is simply @xmath244 .
the total number of droplets per unit volume , @xmath245 , also depends on the fraction @xmath246 because new bubbles can only nucleate in a region that is still in the symmetric phase : @xmath247 this quantity is important for the discussion of structure formation .
the average bubble radius , @xmath248 , is obtained as : @xmath249 the quantities @xmath250 , @xmath245 and @xmath248 are evaluated for the parameter set @xmath227gev , @xmath219gev for different values of the bubble wall velocity @xmath238 .
the results are displayed in figure 5 to 7 .
the completion time @xmath242 depends only weakly on the bubble wall velocity and is roughly @xmath251sec ( see figure 5 ) .
this can be compared to a previous calculation of this quantity @xcite , where a value of roughly @xmath252sec is given .
one sees the effect of enhancing the rate .
the time scale on which the nucleation proceeds is decreased by an order of magnitude . also for the number of bubbles per volume after the completion time
the results can be compared . for three values of the bubble wall velocity , namely @xmath96 , @xmath253 and @xmath254 the corresponding values for @xmath255
can be read of from figure 6 : they are @xmath256 , @xmath257 and @xmath258 . for the same velocities in @xcite the values @xmath259 , @xmath260 and @xmath261 are given ; hence , the number of bubbles per volume is increased substantially .
the average bubble radius @xmath248 for the three velocities under consideration comes out to be the same as in @xcite and is plotted in figure 7 . as the initial radius of the critical bubble is of the order of @xmath262fm it is neglected throughout this calculation .
the presented method to calculate self - consistently the solution of a one - loop field equation has the following essential features : * the complete one - loop determinant and its functional derivative is evaluated .
no expansion in any quantity is needed . in view of the fact that the derivative and heat kernel expansion generate unphysical infrared divergencies order by order
this is an important conceptual advantage of the method presented here . *
coupled with an iterative procedure the numerical computation of the functional derivative of an effective action allows the determination of a self - consistent solution of a full one - loop field equation . *
the divergent and convergent contributions to the effective action can be treated separately .
while the convergent part is calculated numerically the divergent part is evaluated analytically .
so the choice of the regularization and renormalization prescription is free . *
the numerical computation of the functional derivative of an effective action only requires numerical integration of a few differential equations per partial wave .
this can be done with high precision and is not computer time consuming .
in conclusion we find that taking into account the radiative corrections already in the determination of the bounce solution has the following effects : * the width of the temperature interval within that a first order phase transition is possible is decreased and with increasing the higgs boson mass it is shifted towards higher temperatures . *
the transition rate is increased substantially relative to the rate obtained from the standard calculation . * the time scale within that the transition is completed is reduced .
the number of bubbles per volume is increased .
it is a pleasure to thank j. baacke for useful discussions concerning the numerical method and the treatment of the zero modes and for reading the manuscript .
in @xcite the partial wave decomposition of the green s function was done by reduction of the sphaleron system .
an alternative derivation will be given here .
the green s function @xmath98 introduced in ( [ funcder_genform ] ) satisfies the matrix equation @xmath263 the non - zero components of @xmath93 and @xmath94 are given by @xmath264 for the partial wave expansion of the solution @xmath265 of the corresponding homogenous problem we use the following ansatz : @xmath266_j\ , , \quad \varphi_j^2(\mbox{\boldmath$\hat x$\unboldmath } ) : = \left[\mbox{\boldmath$y$\unboldmath}_{l\ l-1\ m } ( \mbox{\boldmath$\hat x$\unboldmath})\right]_j\ , , \quad \varphi_j^3(\mbox{\boldmath$\hat x$\unboldmath } ) : = \left[\mbox{\boldmath$y$\unboldmath}_{l\ l\ m } ( \mbox{\boldmath$\hat x$\unboldmath})\right]_j\,,\\ \varphi_4 ^ 4(\mbox{\boldmath$\hat x$\unboldmath})&:= & \varphi_5 ^ 5(\mbox{\boldmath$\hat x$\unboldmath } ) = \varphi_6 ^ 6(\mbox{\boldmath$\hat x$\unboldmath})= y_{lm}(\mbox{\boldmath$\hat x$\unboldmath})\,.\end{aligned}\ ] ] here @xmath267 means the unit vector in @xmath268direction .
to shorten the notation the @xmath202 and @xmath269dependence of the functions @xmath270 and the @xmath202dependence of the functions @xmath271 is not shown explicitly in the following .
the vector spherical harmonics @xmath272 ( see e. g. @xcite ) are related to the common surface spherical harmonics : @xmath273 for the partial wave @xmath117 one has to consider the following peculiarity : neither the function @xmath274 nor @xmath275 exists , because the normalizing prefactor diverges . scaling @xmath276 with @xmath277 and @xmath278 with @xmath279 , the products @xmath280 and @xmath281
are well defined also at @xmath117 ; they yield zero .
i. e. , at @xmath117 the components @xmath276 and @xmath278 vanish identically .
hence , the coupled @xmath282system reduces to a coupled @xmath283system . in the following the index @xmath202 of the functions
@xmath284 is suppressed .
using the index notation the ansatz ( [ psil ] ) can be formulated as @xmath285 and with @xmath286 the homogeneous equation for @xmath287 takes the form : @xmath288 using @xmath289 we get @xmath290 f_i = g\phi^\prime \zeta_{nk}^i\frac{x_i}{r}\psi_k\,.\ ] ] let @xmath291 be an index that takes values from 1 to 3 .
then we have @xmath292 with the useful relations ( * ? ?
? * s. 103f . )
@xmath293 the right - hand side of ( [ rhsdglpsi ] ) can be written as : @xmath294 therefore , as the vector spherical harmonics are linear independent , ( [ equ : homogeneous ] ) only can be satisfied for @xmath295 being a solution of @xmath296 f_k(r)=0\,,\\ \mbox{with } & ~ & { \mathcal{v}}(\phi)= \left ( \begin{array}{llllll } m_a^2(\phi ) & 0 & 0 & -c_1g\phi^\prime & 0 & 0\\ 0 & m_a^2(\phi ) & 0 & c_0g\phi^\prime & 0 & 0\\ 0 & 0 & m_a^2(\phi ) & 0 & 0 & 0\\ -c_1g\phi^\prime & c_0g\phi^\prime & 0 & m_\varphi^2(\phi , t ) & 0 & 0\\ 0 & 0 & 0 & 0 & m_h^2(\phi , t ) & 0\\ 0 & 0 & 0 & 0 & 0 & m_a^2(\phi)\\ \end{array } \right)\;.\end{aligned}\ ] ] these considerations motivate the following ansatz for the green s function @xmath98 : @xmath297 where the radial green s functions their @xmath202dependence is not displayed explicitly
satisfy @xmath298 g_{km}(r , r^\prime)= -\frac{1}{r^2}\delta(r - r^\prime ) \delta_{nm}\;.\ ] ] in ( [ funcder_genform ] ) one needs to know the following expressions : @xmath299 & = & \sum_{l=0}^{+\infty } \frac{2l+1}{4\pi } \left \ { \frac{3g^2}{4}\phi\left ( g_{11 } + g_{22 } + g_{33}\right ) + \frac{3}{2}\left(\frac{g^2}{2}\phi + 2\lambda_t\phi-3et\right)g_{44 } \right.\nonumber\\ & & \left .
+ 3 ( \lambda_t \phi -et ) g_{55 } -\frac{3g^2}{4}\phi g_{66 } \right \}\;,\\ { \rm tr } \left [ c \ ; \mbox{\boldmath$\zeta$\unboldmath } \mbox{\boldmath$\nabla$\unboldmath } g \right ] & = & \frac{3}{2 } \sum_{l , m } \zeta_{nm}^r \nabla_r g_{ab}\varphi_m^a \varphi_n^{b\ast } = \frac{3}{2 } \sum_{l , m } \left ( \nabla_i g_{4j } \varphi_4 ^ 4 \varphi_i^{j\ast } + \nabla_i g_{j4 } \varphi_i^j \varphi_4^{4\ast}\right)\nonumber\\ & = & \frac{3}{2 } \sum_l \frac{2l+1}{2\pi } \left\ { -c_1 \left ( g_{14}^\prime + \frac{2}{r}g_{14}\right ) + c_0 \left ( g_{24}^\prime+\frac{2}{r}g_{24}\right ) \right\}\;.\end{aligned}\ ] ] obviously the components @xmath300 and @xmath301 satisfy the same differential equation ( see ( [ radial_greenfkt ] ) ) .
hence , they can be combined . due to the different sign of their factors of degeneracy they cancel exactly , except for the s wave .
as mentioned above at @xmath117 the amplitudes @xmath276 and @xmath278 vanish and therefore @xmath302 in this particular partial wave .
hence , the @xmath303 reduces to a @xmath101system , where the fifth channel only contributes at @xmath117 .
the final result is given in ( [ funcder_result ] ) .
10 v. a. kuzmin , v. a. rubakov , and m. e. shaposhnikov , phys .
155b * , 36 ( 1985 ) .
a. g. cohen , d. b. kaplan , and a. e. nelson , ann .
nucl . part .
sci . * 43 * , 27 ( 1993 ) .
s. coleman , phys . rev .
* d15 * , 2929 ( 1977 ) . c. callan and s. coleman , phys
. rev . * d16 * , 1762 ( 1977 ) . j. s. langer , ann
( n.y . ) * 41 * , 108 ( 1967 ) . j. s. langer , ann .
( n.y . ) * 54 * , 258 ( 1969 ) .
i. affleck , phys .
lett . * 46 * , 388 ( 1981 ) .
s. coleman and e. weinberg , phys . rev .
* d7 * , 1888 ( 1973 ) .
m. dine _ et al .
_ , phys . rev . *
d46 * , 550 ( 1992 ) .
d. a. kirzhnits and a. d. linde , ann .
* 101 * , 195 ( 1976 ) .
m. e. shaposhnikov , nucl .
b287 * , 757 ( 1987 ) .
m. e. shaposhnikov , nucl . phys . * b299 * , 797 ( 1988 ) .
g. w. anderson and l. j. hall , phys . rev . * d45 * , 2685 ( 1992 ) .
j. kripfganz , a. laser , and m. g. schmidt , nucl . phys .
* b433 * , 467 ( 1995 ) .
j. kripfganz , a. laser , and m. g. schmidt , z. phys . * c73 * , 353 ( 1997 ) . e. j. weinberg , phys . rev .
* d47 * , 4614 ( 1993 ) .
j. baacke , phys . rev . * d52 * , 6760 ( 1995 ) . j. baacke and a. srig , phys . rev . * d53 * , 4499 ( 1996 ) . j. baacke , h. so , and a. srig , z. phys .
* c63 * , 689 ( 1994 ) .
j. baacke and v. g. kiselev , phys . rev . * d48 * , 5648 ( 1993 ) .
j. baacke and a. srig , z. phys .
* c73 * , 369 ( 1997 ) .
j. baacke , z. phys . *
c47 * , 263 ( 1990 ) .
j. baacke , z. phys .
* c47 * , 619 ( 1990 ) .
j. baacke , z. phys . *
c53 * , 407 ( 1992 ) .
m. grtler , e .-
ilgenfritz , and a. schiller , hep - lat/9704013 ( 1997 ) .
g. d. moore and n. turok , phys . rev . * d55 * , 6538 ( 1997 ) .
m. e. carrington , phys . rev . * d45 * , 2933 ( 1992 ) .
w. buchmller , z. fodor , t. helbig , and d. walliser , ann .
phys . * 234 * , 260 ( 1994 ) .
a. d. linde , rept .
phys . * 42 * , 389 ( 1979 ) .
d. j. gross , r. d. pisarski , and l. g. yaffe , rev .
phys . * 53 * , 43 ( 1981 ) .
s. weinberg , phys .
rev . * d9 * , 3357 ( 1974 ) .
r. d. pisarski , phys .
lett . * 63 * , 1129 ( 1989 ) .
e. braaten and r. d. pisarski , nucl .
* b337 * , 569 ( 1990 ) .
r. r. parwani , phys . rev .
* d45 * , 4695 ( 1992 ) .
p. arnold and o. espinosa , phys
. rev . * d47 * , 3546 ( 1993 ) .
e. w. kolb and m. s. turner , _ the early universe .
_ , _ frontiers in physics ; 69 .
_ ( addison - wesley , redwood city , calif .
[ u.a . ] , 1990 ) . a. h. guth and e. j. weinberg , phys .
* d23 * , 876 ( 1981 ) .
d. bdeker , w. buchmller , z. fodor , and t. helbig , nucl . phys .
* b423 * , 171 ( 1994 ) .
a. r. edmonds , _ drehimpulse in der quantenmechanik _ ( bi - hochschultaschenbcher , mannheim , 1964 ) .
[ cols= " > , > , > , > , > " , ] _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ table 1 : the values of the roll - over temperature @xmath223 and the critical temperature @xmath236 obtained from the high temperature potential @xmath55 and the approximation of the one - loop resummed effective potential @xmath158 .
the top - quark mass is chosen as @xmath219gev .
the higgs boson mass @xmath304 and the temperatures are given in gev .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
* figure 1 : * the expansion of the one - loop correction @xmath305 to the effective action . the lines represent the propagators and the dots the vertex factors given by @xmath306 .
* figure 2 : * comparison of the logarithm of the nucleation rate @xmath0 ( see ( [ nuclrate ] ) ) , i. e. including the one - loop corrections and prefactors , versus the temperature @xmath235 .
@xmath0 has dimension @xmath307 and is calculated in units of @xmath308 .
the squares represent the results we obtain using the profile @xmath164 , which is a solution of ( [ classprofequ ] ) .
they are connected by straight dashed lines .
the dots are the results we obtain using the the self - consistent solution of the full one - loop field equation ( [ fieldequation3 ] ) .
the solid line is the function @xmath309 as defined in ( [ func : f_t ] ) , with @xmath225 and @xmath226 .
the parameter set used here is @xmath227gev and @xmath310=170gev .
* figure 4 : * comparison of the logarithm of the nucleation rate after mapping of the temperature intervals @xmath231 $ ] and @xmath232 $ ] to @xmath233 $ ] . for the parameter set @xmath227gev ( @xmath230gev ) and @xmath219gev the squares ( triangles ) are the results obtained when using the profile @xmath164 and the dots ( diamonds ) are due the self - consistent profile . * figure 5 : * the logarithm of the fraction @xmath250 of space that has been converted to the asymmetric phase in dependence of the time @xmath239 .
the three lines are due to three different values of the bubble wall velocity @xmath238 .
the solid line is obtained for @xmath312 , the short - dashed line for @xmath313 and the long - dashed one for @xmath314 .
the parameter set used here is @xmath227gev and @xmath219gev . * figure 6 : * the total number of bubbles per unit volume @xmath245 in dependence of the time @xmath239 .
the three lines are due to three different values of the bubble wall velocity @xmath238 .
the description is the same as in figure 5 .
* figure 7 : * the average bubble radius @xmath248 in dependence of the time @xmath239 .
the three lines are due to three different values of the bubble wall velocity @xmath238 .
the description is the same as in figure 5 . | in the standard procedure for calculating the decay rate of a metastable vacuum the solution of the classical euclidean equation of motion of the background field is needed . on the other hand radiative corrections have to be taken into account already in the equation of motion .
hence , the latter one has to be the functional derivative of the effective action with respect to the background field .
this is of crucial importance in theories in which the symmetry breaking is due to radiative corrections .
usually the effective potential is considered only , neglecting the corrections due to the derivative terms of the effective action . in this article a bounce solution from an equation of motion which takes into account the full effective action in the one - loop approximation is calculated . a computational method that yields a strict separation of the divergent contributions to the effective action from the convergent ones
is obtained .
this allows a wide freedom in the choice of regularization and renormalization schemes .
the model under consideration is the su(2)-higgs model .
the fluctuations of the complete bosonic sector , i. e. gauge field , higgs and goldstone boson contributions , are taken into account .
the bounce is then self - consistent to one - loop order .
the obtained results for characteristic quantities of the transition as the nucleation rate and the number of nucleated bubbles per volume are compared to other , non - self - consistent approaches . |
statistical physics tells us that systems of many interacting dynamical units collectively exhibit a behavior which is determined by only a few basic dynamical features of the individual units and of the embedding dimension but independent of all other details .
this feature which is specific to critical phenomena , like in continuous phase transitions , is known as universality @xcite .
there is enough empirical evidence that a number of social phenomena are characterized by simple emergent behavior out of the interactions of many individuals . in recent years
, a growing community of researchers have been analyzing large - scale social dynamics to uncover universal patterns and also trying to propose simple microscopic models to describe them , similar to the minimalistic models used in statistical physics .
these studies have revealed interesting patterns and behaviors in social systems , e.g. , in elections @xcite , growth in population @xcite and economy @xcite , income and wealth distributions @xcite , financial markets @xcite , languages @xcite , etc .
( see refs .
@xcite for reviews ) .
academic publications ( papers , books etc . )
form an unique social system consisting of individual publications as entities , containing bibliographic reference to other older publications , and this is commonly termed as _ citation_. the number of citations is a measure of the importance of a publication , and serve as a proxy for the popularity and quality of a publication .
there has already been a plethora of empirical studies on citation data @xcite , specifically on citation distributions @xcite of articles , time evolution of probability distribution of citation @xcite , citations for individuals @xcite and even their dynamics @xcite , and the modeling efforts on the growth and structure of citation networks have produced a huge body literature in network science concerning scale - free networks @xcite , and long - time scientific impact @xcite .
the bibliometric tool of citation analysis is becoming increasingly popular for evaluating the performance of individuals , research groups , institutions as well as countries , the outcomes of which are becoming important in case of offering grants and awards , academic promotions and ranking , as well as jobs in academia , industry and otherwise .
since citations serve as a crucial measure for the importance and impact of a research publication , its precise analysis is extremely important .
annual citations and impact factor of journals are of key interest , primarily from the point of view of journals themselves , and secondarily from the perspective of authors who publish their papers in them . wide distributions of both annual citations and impact factors are quite well studied @xcite .
it is quite usual to find that some publications do better than others due to the inherent heterogeneity in the quality of their content , the gross attention on the field of research , the relevance to future work and so on .
thus different publications gather citations in time at different rates and result in a broad distribution of citations . in 1957 , shockley @xcite claimed that the scientific publication rate is dictated by a lognormal distribution , while a later evidence based on analysis of records for highly cited physicists claim that the citation distribution of individual authors follow a stretched exponential @xcite .
however , an analysis of data from isi claims that the tail of the citation distribution of individual publications decays as a power law with an exponent close to @xmath6 @xcite , while a rigorous analysis of 110 years of data from physical review concluded that most part of the citation distribution fits remarkably well to a lognormal @xcite .
the present consensus lies with the fact that while most part of the distribution does fit to a lognormal , the extreme tail fits to a power law @xcite .
it has been shown earlier @xcite that the distribution of citations @xmath7 to papers within a discipline has a broad distribution , which is universal across broad scientific disciplines , using a relative indicator @xmath8 , where @xmath9 is the average citation within a discipline .
however , it has also been shown later that this universality is not absolutely guaranteed @xcite .
subsequent work on citations @xcite and impact factors @xcite has revealed interesting patterns of universality , some alternative methods have been proposed @xcite and there are also interesting work on citation biases @xcite .
some studies @xcite also report on the possible lack of universality in the citation distribution at the level of articles . a rigorous and detailed study on the citation distributions of papers published in 2005 - 2008 for 500 institutions
@xcite reveals that using the analysis ref .
@xcite , universality condition is not fully satisfied , but the distributions are found to be very similar .
there have also been studies at the level of countries in the same direction @xcite . in this article , we focus on citations received by individual ( i ) academic institutions and ( ii ) academic journals .
we perform the analysis primarily for all articles and reviews , as well as all citable documents . while institutions can vary in their quality of scientific output measurable in terms of total number of publications , total citations etc .
, here we show for the first time that irrespective of the institution s scientific productivity , ranking and research impact , the probability @xmath10 that the number of citations @xmath7 received by a publication is a broad distribution with an universal functional form .
in fact , using a relative indicator @xmath8 , where @xmath9 is the average number of citations to articles published by an institution in a certain year , we show that the effective probability distribution function that an article has @xmath7 citations has the same mathematical form .
we present evidence for the fact that this holds roughly across time for most institutions irrespective of the scientific productivity of the institution considered .
when we carry out a similar analysis on journals , we find similar results .
the scaled distributions fit to a lognormal distribution for most of their range .
again , we find that these features roughly hold across time and across journals within the same class .
the largest citations for academic institutions as well as the journals seem to fit well to a power law .
we also present evidence that each of these sampled groups
institutions , and journals are distinct with the absolute measure of inequality as computed from their distribution functions , with high absolute inequality for both these sets , the gini coefficients being around @xmath0 and @xmath1 for institutions and journals respectively .
we also find that the top @xmath2% of the articles fetch about @xmath3% of the total citations for institutions and the top @xmath4% of the articles fetch about @xmath5% of the total citations for journals .
we collected data from 42 academic institutions across the world .
institutions were selected such that they produce considerable amount of papers ( typically 200 or more ) so that reasonable statistics could be obtained . however , there were exceptions for certain years for particular institutions .
all papers published with at least one author with the institution mentioned as affiliation were collected .
this was done for 4 years 1980 , 1990 , 2000 , 2010 .
we also selected 30 popular academic journals across physics , chemistry , biology and medicine .
however , for some journals , only 3 years of data could be collected , since they were launched after 1980 .
the citable papers considered in this study are articles and reviews , although we compare the results with the same analysis done on all citable documents .
we study the data of number of citations to publications from different years , from isi web of science @xcite for several ( i ) academic institutions ( research institutes and universities ) and ( ii ) popular journals .
it is to be noted that citations to individual publications arrive from any publication indexed in isi web of science and does not mean only internal citations within the journal in which it is published .
we analyzed data of science publications from 42 academic institutions and 30 popular journals .
we recorded the data for the number of papers published , the total number of citations to each of the publications , for a few years ( 1980 , 1990 , 2000 , 2010 for most cases ) . since citations grow with time , we have studied publications which are at least 4 years old ( from 2010 ) or more ( 1980 , 1990 , 2000 ) to rule out any role of transients .
we also collect data from academic institutions and journals which have a comparatively large number of publications , so as to produce good statistics , and minimize the effects of aberration that can result from fluctuations of the quantities measured from small data sets .
we collected citation data until date for all articles and reviews from a particular year ( e.g. 1980 , 1990 , 2000 , 2010 ) . for each year , the probability distribution @xmath10 of citations
@xmath7 for an academic institution was observed to be broad .
for instance , fig .
[ fig : inst_all_1990]a shows the plot of @xmath10 vs. @xmath7 for various institutions for publications from 1990 .
we rescaled the absolute value of citation for each year by the average number of citations per publication @xmath9 , and plotted this quantity @xmath8 against the adjusted probability @xmath11 ( fig .
[ fig : inst_all_1990]b ) ( see similar plots for 1980 and 2000 in fig . s1 of si ) .
we remarkably find that the distributions collapse into an universal curve irrespective of the wide variation in the academic output of the different institutions .
the scaling collapse is good for more than 3 decades of data and over 5 orders of magnitude .
the average number of papers , total citations and the average number of citations per publication are shown in table s3 .
the rescaled curves fit well to a lognormal @xmath12 \label{eq : ln}\ ] ] with @xmath13 with @xmath14 , for a considerable range of the distribution .
however , if one fits a lognormal distribution to individual sets , the range of parameters are quite narrow , @xmath15 lies in the range @xmath16 to @xmath17 , while @xmath18 lies in the range @xmath19 to @xmath20 .
the fitting were performed using a least square fitting routine .
for lowest values of the abscissa , seems to follow @xmath21 or slowly growing , as @xmath22 .
however , the largest citations deviate from the lognormal fit and are better described according to @xmath23 , with @xmath24 ( see si table s6 for exponents for other years ) .
the power law exponent has been estimated using the maximum likelihood estimate method ( mle ) @xcite . in order to investigate
if the distributions @xmath10 for different institutes vary with time , we plot the same for each institution for several years .
the rescaled plots show scaling collapse indicating that although the average citations vary over years , the form of the distribution function remain roughly invariant , when scaled with the average number of citations .
[ fig : inst_years ] shows the plot for 1990 . to check if this also holds for time - aggregated data , we collected citations for all papers published during the period 2001 - 2005 for the same set of institutions , and repeated the above analysis ( see si . fig .
s2 ) .
we collected citation data until date for all articles and reviews in individual journals for several years ( e.g. 1980 , 1990 , 2000 , 2010 etc . ) . for each year , the probability distribution @xmath10 of citations
@xmath7 was again observed to be broad . as in the case of institutions , we plotted @xmath8 against the adjusted probability @xmath11 ( fig .
[ fig : joun_all ] ) . for a particular journal
, it is observed that the curves follow similar distributions over years although the average number of papers , total citations and hence the average number of citations vary ( see table s4 in si for details ) .
further , we plot the same quantity for a particular year for different journals ( see fig . [
fig : jour_class_1990 ] for 1990 ) , and find that the curves roughly collapse into an single curve irrespective of the wide variation in the output of the different journals .
the bulk of the rescaled distribution fits well to a lognormal form with @xmath25 and @xmath26 , as was observed in case of institutions , while the largest citations fit better to a power law @xmath23 , with @xmath27 ( see si table s6 for exponents for other years ) . to check if this also holds for time - aggregated data , we collected citations for all papers published during the period 2001 - 2005 for the same set of journals and repeated the analysis ( see si . fig .
however , we observed that if we consider all citable documents , two distinct classes of journals emerge according to the shape of the distributions to which the curves collapse .
the first group is a _
general _ class , for which most of the distribution fits well to a lognormal function even quite well for the lowest values of the abscissa .
this is similar to what is observed for all journals if we consider only articles and reviews .
the other group , which we call the _ elite _ class ( si fig .
s5 ) is also broadly distributed but has a distinct and faster monotonic decay compared to the _ general _ class , where @xmath28 , i.e. , @xmath29 with @xmath30 .
this divergence at the lowest values of citations also indicate that the _ elite _ journals have a larger proportion of publications with less number of citations although their average number of citations @xmath9 is larger than those for the general class .
however , for both the above classes , the largest citations still follow a power law @xmath23 , with @xmath31 for the _ general _ class and @xmath32 for the _ elite _ class .
we reason for such a behavior in the _ elite _ journal class is because of a large fraction @xmath33 of uncited documents .
if we consider only articles and reviews , @xmath33 is usually 2 - 10% .
considering all citable documents , this fraction does not change appreciably for the _ elite _ class of journals , and can be anything in the range 25 - 80% ( see values of @xmath33 in si table s5 ) , and are primarily in the category of news , correspondence , editorials etc .
such documents in the _ general _ class is either absent or are very few .
we are able to find at least 7 journals ( see si fig .
s5 and table s5 ) in the _ elite _ class while most of others belong to the _ general _ class .
the power law tail in all distribution suggests that the mechanism behind the popularity of the very highly cited papers is a ` rich gets richer ' phenomena @xcite ( see fig . s3 of si for 1980 and 2000 ) . following ref .
@xcite , we rank all articles belonging to different institutions according to @xmath7 and @xmath34 .
we then compute the percentage of publications of each institution that appear in the top @xmath35% of the global rank .
the percentage for each should be around @xmath35% with small fluctuations if the ranking is good enough .
the same is performed for journals .
when ranking is done according to unnormalized citations @xmath7 then the frequency distribution of @xmath35% of papers is wide .
however , if the ranking is done according to normalized citations @xmath34 , then the frequency distribution is much narrow .
for example , we show the results for institutions and journals in fig . [
fig : top10 ] if @xmath36% .
assuming that articles are uniformly distributed on the rank axis , the expected average bin height must be z% with a standard deviation given by @xmath37 where @xmath38 is the number of entries ( institutions or journals ) and @xmath39 is the number of papers for the @xmath40-th institution or journal . for institutions , when the ranking is done according to @xmath34 we observe that the theoretically calculated value ( from above equation ) of @xmath41 is @xmath42 compared to @xmath43 as computed directly from the fitting , while if the ranking was done according to @xmath7 , @xmath41 is @xmath44 .
similarly for journals , @xmath41 computed from the above equation is @xmath45 compared to @xmath46 as computed from the fitting , while if the ranking was done according to @xmath7 , @xmath41 is @xmath47 .
this indicates that @xmath34 is indeed an unbiased indicator , as seen earlier @xcite .
we calculate absolute measures of inequality like the commonly used gini index @xcite as well as the @xmath48-index @xcite which tells us that the top cited @xmath49 fraction of papers have @xmath48 fraction of citations , and we report in si tables .
s3 , s4 , s5 . for academic institutions , gini index @xmath50 and @xmath51 , which means around
@xmath52 citations come from the top @xmath53 papers . for journals , @xmath54 , @xmath55 which means about
@xmath56 citations come from the top @xmath57 papers .
we further note that gini and @xmath48 indices fluctuate less around respective mean values @xmath58 and @xmath59 as the number of articles and number of citations become large ( fig .
[ fig : gk_conv ] ) . for academic institutions ,
the values are @xmath60 for gini and @xmath61 .
for journals , the values are @xmath62 and @xmath63 .
in this article we analyze whether the citations to science publications from academic institutions ( universities , research institutes etc . ) as well as journals are distributed according to some universal function when rescaled by the average number of citations . for institutions , it seems to fit roughly to a log - normal function .
the largest citations , however , deviate from the lognormal fit , and follow a power law decay .
this rough universality claim is an interesting feature , since for institutions , the quality of scientific output measurable in terms of the total number of publications , total citations etc .
vary widely across the world as well as in time .
nevertheless , the way in which the number of papers with a certain number of citations is distributed is quite similar @xcite , seems to be quite independent of the quality of production / output of the academic institution .
although there has been claims that the form of the distribution of citations for different scientific disciplines are the same @xcite , albeit deviations @xcite , it is also true that each discipline is characterized by a typical average number of citations @xmath64 . as a matter of fact , that different institutions have a varying strength of publication contribution towards different disciplines makes the issue of obtaining a universal function for the resulting ( effective ) distribution of citations ( for the institution ) quite nontrivial . in other words ,
different academic institutions have a variety in the strength of their academic output , in terms of variation of representations across different disciplines and the amount of citations gathered .
this does not necessarily guarantee that the universality which has been already reported across disciplines @xcite will still hold when one looks at data from different institutions , rest aside the counter claims about lack of universal character @xcite for citation distribution across distinct disciplines .
there are already critical studies on the citation distribution of universities @xcite using larger data sets , which raises issues on the nature of universality .
we observe similar features for academic journals the bulk of the probability distribution fitting reasonably well to a lognormal while the highest cited papers seem to fit well to a power law decay with a similar exponent ( @xmath65 ) .
we note that the exponents are consistently less than @xmath6 , the exponent of the full citation distribution @xcite , which is due to the fact that our data are very small subsets , which fall short of catching the correct statistical behavior of all of the highest cited papers .
our results indicate that dividing citation counts by their average indeed helps to get closer to universal citation distributions .
however , the results also indicate that , even after such a rescaling , @xmath41 are substantially larger than the theoretical values
@xmath42 compared to @xmath43 while it is @xmath44 for unscaled data for institutions and @xmath45 compared to @xmath66 while it is @xmath47 for unscaled data for journals .
this indicates that the universality is not very strong , and holds only in an approximate sense .
@xcite shows similar evidence for institutions , claiming the absence of universality but pointing out the similarity between the distributions .
another previous study @xcite on different fields of science also reported that this universality claim does not hold very well for all fields .
we further note that the inequality in the distribution of citations of institutions and journals differ quantitatively .
as the number of papers and citations increase , the absolute measures of inequality like gini and @xmath48 indices seem to converge to different values for the above two sets .
the values of gini index are @xmath0 and @xmath1 for institutions and journals respectively .
the @xmath48 index values suggest that the top @xmath2% of the articles hold about @xmath3% of the total citations for institutions and the top @xmath4% of the articles hold about @xmath5% of the total citations for journals .
the authors thank s. biswas for discussions , j .- i .
inoue , s. redner and p. sen for useful comments .
a.c . and b.k.c .
acknowledge support from b.k.c.s j. c. bose fellowship research grant .
inoue ji , ghosh a , chatterjee a , chakrabarti bk ( 2015 ) measuring social inequality with quantitative methodology : analytical estimates and empirical data analysis by gini and @xmath48 indices .
physica a 429 : 184 - 204 . | citations measure the importance of a publication , and may serve as a proxy for its popularity and quality of its contents . here
we study the distributions of citations to publications from individual academic institutions for a single year .
the average number of citations have large variations between different institutions across the world , but the probability distributions of citations for individual institutions can be rescaled to a common form by scaling the citations by the average number of citations for that institution .
we find this feature seem to be universal for a broad selection of institutions irrespective of the average number of citations per article .
a similar analysis for citations to publications in a particular journal in a single year reveals similar results .
we find high absolute inequality for both these sets , gini coefficients being around @xmath0 and @xmath1 for institutions and journals respectively .
we also find that the top @xmath2% of the articles hold about @xmath3% of the total citations for institutions and the top @xmath4% of the articles hold about @xmath5% of the total citations for journals . |
let @xmath2 be probability spaces and @xmath3 be a fixed measurable function .
the monge s optimal transportation problem is the minimization of the following functional @xmath4 over all the borel maps @xmath5 which pushes forward @xmath6 to @xmath7 : @xmath8 .
maps @xmath9 which achieve the infimum above are called optimal maps . in this paper
, we will only consider the case when @xmath10 is a manifold . in 1942 , kantorovich studied a relaxed version of the monge s problem in his famous paper @xcite .
however , a huge step toward solving the original problem is not achieved until a decade ago by brenier . in @xcite ,
brenier proved the existence and uniqueness of optimal map in the case where @xmath11 and the cost function @xmath12 is given @xmath13 .
later , this is generalized to the case of a closed riemannian manifold @xmath14 with cost given by the square of the riemannian distance @xmath15 by mccann @xcite .
recently , bernard and buffoni @xcite generalized this further to the case where the cost @xmath12 is the action associated to a lagrangian function @xmath16 on a compact manifold @xmath14 .
more precisely , @xmath17 where the infimum is taken over all curves joining the points @xmath18 and @xmath19 , and the lagrangian @xmath20 is fibrewise strictly convex with superlinear growth . in this paper
, we consider costs similar to ( [ lagcost ] ) .
however , instead of minimizing among all curves , the infimum is taken over a subcollection of curves , called admissible paths .
these paths are given by a control system and the corresponding cost function is called the optimal control cost .
roughly speaking , a control system is a smooth fibrewise mapping of a locally trivial bundle over @xmath14 into @xmath21 .
locally , such a mapping has a form @xmath22 , where @xmath23 , and @xmath24 is a typical fiber .
we assume that @xmath24 is a closed subset of a euclidean space .
admissible controls are measurable bounded maps from @xmath25 $ ] to @xmath24 , while admissible paths are lipschitz curves which satisfy the equation @xmath26 where @xmath27 is an admissible control .
let @xmath28 be a lagrangian , then the corresponding cost @xmath12 is given by @xmath29 where the infimum is taken over all admissible pairs @xmath30\to
m\times u$ ] such that @xmath31 . in the interesting cases ,
the dimension of @xmath24 is essentially smaller than that of @xmath14 and , nevertheless , any two points of @xmath14 can be connected by an optimal admissible path . in other words ,
the control system works as a nonholonomic constraint .
shortage of admissible velocities does not allow to recover an optimal path from its initial point and initial velocity and the euler
lagrange description of the extremals does not work well . on the other hand ,
hamiltonian approach remains efficient according to the pontryagin maximum principle .
another problem is appearance of so called abnormal extremals ( singularities of the space of admissible paths ) which we are obliged to fight with . in sections 2 and 3 , we will recall some basic notions in optimal control theory and the theory of optimal mass transportation which are necessary for this paper . in section 4 , by using some standard argument in the theory of optimal mass transportation and the pontryagin maximum principle in optimal control theory
, we show the existence and uniqueness of optimal map under some regularity assumptions ( theorem [ main ] ) .
all these conditions are mild except the lipschitz continuity of the cost function .
however , this is well - known in all of the above cases mentioned .
so , the theorem generalizes the work in @xcite . in section 5
, we study the lipschitz continuity of the cost function . if abnormal minimizers are absent ,
then the cost is not only lipschitz but even semi - concave ( see @xcite ) .
unfortunately , abnormal minimizers are unavoidable in many interesting problems and , in particular , in all subriemannian problems .
it happens however that not all abnormal minimizers are dangerous . to keep the lipschitz property of the cost , ( though not the semi - concavity ) it
is sufficient that the , so called , sharp abnormal minimizers are absent .
sharp paths are essentially singularities of the space of admissible paths whose neighborhoods in the second order approximations are contained in quadrics with a finite morse index .
geometric control theory provides simple effective conditions of the sharpness ( see , for instance , @xcite ) .
these conditions allow us to prove lipschitz continuity for a large class of optimal control cost . hence , proving the existence and uniqueness of optimal map of the corresponding monge s problem ( theorem [ maintransport ] ) . in section 6 ,
we apply the above results to some subriemannian manifolds , where the cost function is given by the square of the subriemannian distance ( see section 6 for the basic notions in subriemannian geometry ) . in the case of a subriemannian manifold ,
all the mild regularity assumptions are satisfied . using the result in @xcite mentioned above ( proposition [ goh ] )
, lipschitz continuity of the cost can be easily proven in the case of a step 2 distribution ( corollary [ lip2 ] ) .
hence , proving existence and uniqueness of optimal map .
this generalizes the corresponding result by ambrosio and rigot @xcite on the heisenberg group . in section 7 and 8 ,
we show some properties of the optimal plan when abnormal minimizers are present . in section 7
, we consider flows whose trajectories are strictly abnormal minimizers .
we show that these flows can not be an optimal plan for all nice " initial measures if the cost is continuous . on the contrary , in section 8
, we show that these flows are indeed optimal for an important class of problems with discontinuous cost . in section 9
, we study two examples on grushin plane for which the results in section 3 and 4 apply .
in this section , we recall some notions from optimal control theory . see @xcite , @xcite for detail .
let @xmath14 be a smooth manifold and let @xmath24 be a closed subset in @xmath32 which is called the control set .
let @xmath33 be a lipschitz continuous function such that @xmath34 is a smooth vector field for each point @xmath35 in the control set .
assume that the function @xmath36 is continuous .
curves @xmath37\to u$ ] in the control set @xmath24 which are locally bounded and measurable are called admissible controls .
a control system is the following ordinary differential equations with parameters varying over all admissible controls .
@xmath38 the solutions @xmath39 to the above control system are called admissible paths and @xmath40 are called admissible pairs . by classical theory of ordinary differential equations , a unique solution to the system ( [ control ] ) exists locally for almost all time @xmath41 .
moreover , the resulting local flow is smooth in the space variable @xmath18 and lipschitz in the time variable @xmath41 .
the control system is complete if the flows of all control vector fields exist globally .
let @xmath42 and @xmath43 be two points on the manifold @xmath14 .
denote by @xmath44 the set of all admissible pairs @xmath45 for which the corresponding admissible paths @xmath46 start at the point @xmath42 . and denote by @xmath47 those pairs in @xmath44 whose admissible paths end at @xmath43 .
a control system is called controllable if the set @xmath47 is always nonempty for any pair of points @xmath42 and @xmath43 on the manifold .
let @xmath28 be a smooth function , called lagrangian , and defined the cost function corresponding to this lagrangian as follow : @xmath48 the cost function defined above is said to be complete if given any pairs of points @xmath49 , there exists an admissible pair which achieves the infimum above and the corresponding admissible path starts from @xmath42 and ends at @xmath43 .
the infimum of the problem in ( [ cost ] ) can be equivalently characterized by taking infimum over all admissible controls @xmath27 such that the corresponding admissible paths start at the point @xmath43 , end at the point @xmath42 of the manifold and satisfy the following control system @xmath50 this point will become important for the later discussion .
consider the following minimization problem , commonly known as the bolza problem : [ bolza ] @xmath51 next , we present an elementary version of the pontryagin maximum principle which we prove in the appendix for the convenience of readers .
let @xmath52 be the cotangent bundle projection .
for each point @xmath35 in the control set @xmath24 , define the corresponding hamiltonian function @xmath53 by @xmath54 if @xmath55 is a function on the cotangent bundle , we denote its hamiltonian vector field by @xmath56 . [ pmpb](pontryagin maximum principle for bolza problem ) let @xmath57 be an admissible pair which achieves the infimum in problem [ bolza ] .
assume that the function @xmath58 in problem [ bolza ] is sub - differentiable at the point @xmath59 .
then , for each @xmath60 in the sub - differential @xmath61 of @xmath58 , there exists a lipschitz path @xmath62\to t^*m$ ] which satisfies the following for almost all time @xmath41 in the interval @xmath25 $ ] : @xmath63 [ generalcontrol ] let @xmath64 be a distribution on a @xmath65-dimensional manifold @xmath14 .
that is , for each point @xmath18 in the manifold @xmath14 , it smoothly assigns a vector subspace @xmath66 of the tangent space @xmath67 .
assume that the distribution @xmath68 is trivializable , i.e. there exists a system of vector fields @xmath69 which span @xmath68 at every point : @xmath70 .
consider the following control system : @xmath71 with initial condition @xmath72 and final condition @xmath73 . recall that we denote by @xmath74 the set of all admissible pair @xmath45 such that the admissible path @xmath46 satisfies @xmath72 and @xmath73
let @xmath12 be the cost given by @xmath75 if @xmath76 is equal to the dimension @xmath65 of the manifold @xmath14 , then the distribution @xmath68 is the same as the tangent bundle @xmath21 of @xmath14 and the admissible paths of the control system ( [ simplecontrol ] ) are all the paths on @xmath14
. it also defines a riemannian metric on @xmath14 by declaring that the vector fields @xmath77 are orthonormal everywhere .
the cost function @xmath12 is the square of the riemannian distance @xmath1 : @xmath78 .
and the minimizers of this system correspond to the length minimizing geodesics on @xmath14 .
however , this does not work for distributions which are not trivializable .
( for instance , any vector field on the @xmath79-sphere has a zero . ) to overcome this difficulty , we can modify the general definition of control system in the following way .
let @xmath80 be a locally trivial bundle on @xmath14 with bundle projection @xmath81 and let @xmath82 be fibre preserving map .
i.e. @xmath83 . the control system corresponding to the map @xmath84 is given by @xmath85 the admissible pairs are locally bounded measurable paths @xmath86\to v$ ] in @xmath80 such that its projection to the manifold @xmath14 is a lipschitz path : @xmath87 is lipschitz .
if we let @xmath80 be the trivial bundle @xmath88 , we get back the system ( [ control ] ) .
if a lagrangian @xmath89 is fixed , then the corresponding cost function @xmath12 is defined by @xmath90 where the infimum is taken over all admissible pair @xmath86\to v$ ] such that the corresponding admissible path @xmath87 satisfies @xmath72 and @xmath73 .
let @xmath91 be a riemannian metric on the manifold @xmath14 .
if @xmath80 is the tangent bundle @xmath21 of @xmath14 , the map @xmath84 is the identity map and the lagrangian @xmath89 given by @xmath92 , then the cost function @xmath12 is equal to the square of the riemannian distance .
if @xmath93 , then the admissible paths of the control system ( [ simplecontrol ] ) are paths tangent to the distribution @xmath68 .
similar to the riemannian case , the control system defines a subriemannian metric @xmath94 .
( see section 6 for the basics on subriemannian geometry ) and the cost ( [ simplecost ] ) is the square of the subriemannian distance @xmath95 : @xmath96 . for
general distributions @xmath68 which are not trivializable , consider the general control system ( [ generalcontrol ] ) with @xmath97 .
and @xmath98 is the inclusion map .
if the lagrangian @xmath20 is defined by @xmath99 , then the cost is again the square of the subriemannian distance . in this paper ( except in section 8)
, we consider the control systems of the form ( [ control ] ) in order to avoid heavy notations .
all the results have easy generalization to more general intrinsically defined systems just introduced .
the theory of optimal mass transportation is about moving one mass to another that minimizes certain cost .
more precisely , let @xmath14 be a manifold and consider a function @xmath100 , called the cost function .
let @xmath6 and @xmath7 be two borel probability measures on the manifold @xmath14 , then the optimal mass transportation is the following problem : [ optimal ] find a borel map which achieves the following infimum among all borel maps @xmath101 that pushes the probability measure @xmath6 forward to @xmath7 @xmath102 here , we recall that the push forward @xmath103 is defined by @xmath104 for all borel set @xmath105 in @xmath14 . in many cases ,
such a problem admits solution which is unique ( up to measure zero ) , assuming absolute continuity of the measure @xmath6 with respect to the lebesgue measure .
this unique solution to ( [ optimal ] ) is called the optimal map or the brenier map .
the first optimal map was found by brenier in @xcite in the case where the manifold is @xmath106 and the cost was @xmath13 .
later , it was generalized to arbitrary closed , connected riemannian manifolds in @xcite with cost given by square of the riemannian distance .
the case for the heisenberg group with the cost given by @xmath0 was done in @xcite , where @xmath1 is the subriemannian distance or the gauge distance . in @xcite ,
a much general cost given by the action associated to a lagrangian function @xmath16 on a compact manifold @xmath14 was considered .
more precisely , @xmath107 where the infimum is taken over all curves joining the points @xmath18 and @xmath19 . existence and uniqueness of optimal map with the cost given by ( [ lag ] ) was shown under the following assumptions : * the lagrangian @xmath20 is fibrewise strictly convex , i.e. the map restriction of @xmath20 to the tangent space @xmath67 is strictly convex for each fixed @xmath18 in the manifold @xmath14 .
* @xmath20 has superlinear growth , i.e. @xmath108 as @xmath109 .
* the cost @xmath12 is complete , i.e. the infimum ( [ lag ] ) is always achieved by some @xmath110 smooth paths .
recently , the compactness assumption on the manifold or on the measures was eliminated @xcite . in this paper , we consider a connected manifold @xmath14 without boundary and the cost function @xmath12 is given by ( [ cost ] ) .
consider the following relaxed version of problem [ optimal ] , called kantorovich reformulation .
let @xmath111 and @xmath112 be the projection onto the first and the second component respectively .
let @xmath113 be the set of all joint measures @xmath114 on the product manifold @xmath115 with marginals @xmath6 and @xmath7 : @xmath116 and @xmath117 .
[ kanre ] @xmath118 if @xmath9 is an optimal map in the problem in ( [ optimal ] ) , then @xmath119 is a joint measure in the set @xmath113 . therefore , problem [ kanre ] is a relaxation of the problem in ( [ optimal ] ) . before we proceed into the existence proof of optimal map ,
let us look at the following dual problem of kantorovich .
see @xcite for history and the importance of this dual problem to optimal transport .
let @xmath12 be a cost function and let @xmath58 be a function on the manifold @xmath14 .
the @xmath120-transform of the function @xmath58 is the function @xmath121 given by @xmath122.\ ] ] similarly , the @xmath123-transform of the function @xmath58 is given by @xmath124.\ ] ] the function @xmath58 is a @xmath12-concave function if it satisfies @xmath125 . let @xmath126 be the set of all pairs of functions @xmath127 on the manifold such that @xmath128 and @xmath129 are in @xmath130 and @xmath131 respectively , and @xmath132 for all @xmath133 .
the dual problem of kantorovich is the following maximization problem : [ dual ] @xmath134 the existence of solution to the above problem is well - known .
see @xcite and @xcite for the proof .
[ dualexist ] assume that there exists two functions @xmath120 and @xmath123 such that @xmath120 is @xmath6-measurable , @xmath123 is @xmath7-measurable and the cost function @xmath12 satisfies @xmath135 for all @xmath136 in @xmath115 .
if @xmath12 is also continuous , bounded below and @xmath137 , then there exists a @xmath12-concave function @xmath58 such that the function @xmath58 is in @xmath130 , its @xmath120-transform @xmath121 is in @xmath131 and the pair @xmath138 achieves the supremum in problem [ dual ] .
the following theorem on the regularity of the dual pair above is also well - known .
[ dualregular ] assume that the cost @xmath139 is continuous , bounded below and the measures @xmath6 and @xmath7 are compactly supported .
then the functions @xmath58 and @xmath121 are upper semicontinuous .
if the function @xmath140 is also locally lipschitz on an set @xmath141 and the lipschitz constant is independent of @xmath19 locally , then @xmath58 can be chosen to be locally lipschitz on @xmath141 .
fix @xmath142 .
since @xmath143 $ ] , there exists @xmath19 such that @xmath144 . also , we have @xmath145 for any @xmath146 in @xmath14 .
so , combining the above equations and continuity of the cost @xmath12 , we have @xmath147 for any @xmath146 close enough to @xmath18 .
therefore , @xmath58 is upper semicontinuous .
let @xmath148 be a compact set containing the support of the measures @xmath6 and @xmath7 .
let @xmath149 then the pair @xmath150 achieves the maximum in problem [ dual ] .
let @xmath151 , then the pair @xmath152 also achieves the maximum . by definition of @xmath153 ,
we have @xmath154 $ ] . by an argument the same as the proof of upper semicontinuity , for any @xmath18 and @xmath146 in the compact subset @xmath155 of @xmath141 , we can find @xmath19 in @xmath148 such that @xmath156 by the assumption of the cost @xmath12 , the above inequality becomes @xmath157 for some constant @xmath158 which is independent of @xmath18 on @xmath155 . by switching the roles of @xmath18 and @xmath146
, the result follows .
the following theorem about minimizers of the problem [ kanre ] is well - known .
see , for instance , @xcite .
[ existjoint ] if we make the same assumption as in theorem [ dualexist ] , then problem [ kanre ] admits a minimizer .
moreover , the joint measure @xmath114 in the set @xmath113 achieve the infimum in problem [ kanre ] if and only if @xmath114 is concentrated on the set @xmath159
in this section , we show that monge s problem with cost given by an optimal control cost ( [ controlcost ] ) can be solved under certain regularity assumptions .
let @xmath55 be the function defined by @xmath160 if @xmath161 is well - defined and @xmath110 , then we denote its hamiltonian vector field by @xmath56 and let @xmath162 be its flow .
let @xmath58 be the function defined in theorem [ dualexist ] which is lipschitz for @xmath6-almost all @xmath18 .
consider the map @xmath163\to m$ ] defined by @xmath164 .
[ main ] the map @xmath165 is the unique ( up to @xmath6-measure zero ) optimal map to the problem ( [ optimal ] ) with cost @xmath12 given by ( [ cost ] ) under the following assumptions : 1 . the measures @xmath6 and @xmath7 are compactly supported and absolutely continuous with respect to the lebesgue measure
@xmath12 is bounded below and @xmath139 is also locally lipschitz in the @xmath18 variable and the lipschitz constant is independent of @xmath19 locally .
the cost @xmath12 is complete , i.e. given any pairs of points @xmath49 in the manifold @xmath14 , there exists an admissible pair @xmath45 such that the pair achieves the infimum in ( [ cost ] ) , @xmath27 is locally bounded measurable , @xmath166 and @xmath167 .
the hamiltonian function @xmath161 defined in ( [ maxham ] ) is well - defined and @xmath110 .
the hamiltonian vector field @xmath56 is complete , i.e. global flow exists .
the rest of this section is devoted to the proof of theorem [ main ] .
let @xmath168 be the set of all admissible pairs such that the corresponding admissible paths @xmath46 starts from the point @xmath19 : @xmath169 and satisfies the following control system : @xmath170 let @xmath171 be the set of all those pairs in @xmath168 such that the corresponding admissible paths @xmath46 end at the point @xmath18 : @xmath172 .
first , we have the following simple observation . [ dualmin ]
let @xmath18 be a point which achieves the infimum @xmath173 and let @xmath174 be an admissible pair in @xmath171 such that the corresponding admissible path @xmath175 minimizes the cost given by @xmath176 then @xmath57 achieves the following infimum @xmath177 if @xmath178 , then @xmath179 achieves the following infimum @xmath180 where @xmath181 denotes the set of all admissible pairs @xmath45 satisfying the following control system : @xmath182 let @xmath183 be as in the above proposition and let @xmath184 .
let @xmath185 be given by @xmath186 .
the following is a consequence of theorem [ pmpb ] .
[ apppmp ] let @xmath175 be a curve that achieves the infimum in ( [ dualmin1 ] ) and let @xmath178 .
assume that @xmath60 is contained in the subdifferential of the function @xmath58 at the point @xmath187 , then there exists a lipschitz curve @xmath188\to t^*m$ ] in the cotangent bundle such that the followings are true for almost all time @xmath41 in the interval @xmath25 $ ] : @xmath189 by theorem [ pmpb ] , there exists a curve @xmath62\to t^*m$ ] in the cotangent bundle @xmath190 such that @xmath191 where @xmath192 .
let @xmath193 and @xmath184 , then the equations above become @xmath194 assume that the hamiltonian function @xmath55 defined by @xmath195 is well - defined and @xmath110 .
let @xmath56 be the hamiltonian vector field of the function @xmath161 and let @xmath196 be its flow .
the function @xmath58 defined in theorem [ dualexist ] is lipschitz and so it is differentiable almost everywhere by rademacher theorem .
moreover , the map @xmath197 is measurable and locally bounded .
so , if we let @xmath163\to m$ ] be the map defined by @xmath164 , then the map @xmath198 is a borel map .
[ dualunique ] under the assumptions of theorem [ main ] , the following is true for @xmath6-almost all @xmath18 : given a point @xmath18 in the support of @xmath6 , there exists a unique point @xmath19 such that @xmath199 moreover , the points @xmath18 and @xmath19 are related by @xmath200 .
we first claim that the infimum @xmath201 $ ] is achieved for @xmath6 almost all @xmath18 . indeed , by assumption , we have @xmath202 for all @xmath136 in @xmath115 .
also , let @xmath114 be the measure defined in theorem [ existjoint ] , then @xmath203 for @xmath114-almost everywhere .
since the first marginal of the measure @xmath114 is @xmath6 , the following is true for @xmath6 almost all @xmath18 : given a point @xmath18 in the manifold @xmath14 , there exists @xmath19 in @xmath14 such that @xmath203 .
this proves the claim .
fix a point @xmath18 for which the infimum @xmath204 $ ] is achieved and let @xmath19 be the point which achieves the infimum . by the proof of the above claim , @xmath18 achieves the infimum @xmath205 $ ] .
therefore , by completeness of the cost @xmath12 and proposition [ dualmin ] , there exists an admissible path @xmath179 such that @xmath206 , @xmath207 and @xmath179 achieves the infimum ( [ dualmin2 ] ) . since @xmath58 is lipschitz on a bounded open set @xmath141 containing the support of @xmath6 and @xmath7 , it is almost everywhere differentiable on @xmath141 by rademacher theorem . since @xmath6 is absolutely continuous with respect to the lebesgue measure , @xmath58 is also differentiable @xmath6-almost everywhere . by theorem [ apppmp ] , for @xmath6-almost all @xmath18 , there exists a curve @xmath188\to t^*m$ ] in the cotangent bundle @xmath190 such that @xmath208 where @xmath209 is the function on the cotangent bundle @xmath190 given by @xmath210 . by the definition of @xmath161 , we have @xmath211 .
but , we also have @xmath212 for all @xmath213 .
since both @xmath161 and @xmath209 are @xmath110 , we have @xmath214 .
hence , @xmath215 for almost all @xmath41 .
the result follows from uniqueness of solution to ode . the rest of the arguments for the existence and uniqueness of optimal map follow from theorem [ existjoint ] .
as mentioned above , problem [ kanre ] is a relaxation of problem [ optimal ] .
we can recover the later from the former by restricting the minimization to joint measures of the form @xmath119 , where @xmath9 is any borel map pushing forward @xmath6 to @xmath7 .
therefore , the results follow from theorem [ existjoint ] and proposition [ dualunique ] .
in theorem [ main ] , we prove existence and uniqueness of optimal maps under certain regularity conditions on the cost .
most of the conditions in the theorem are easy to verify except condition ( 2 ) and ( 3 ) . in this section
, we will give simple conditions which guarantee this regularity .
this includes the completeness and the lipschitz regularity of the cost .
first , we recall some basic notions in the geometry of optimal control problems , see @xcite and reference therein for details .
fix a point @xmath42 on the manifold @xmath14 and assume that the control set @xmath24 is @xmath216 .
denote by @xmath44 the set of all admissible pairs @xmath45 such that the corresponding admissible paths @xmath46 starts at @xmath42 .
moreover , we assume that the control system is of the following form : @xmath217 where @xmath218 and @xmath219 are fixed smooth vector fields on the manifold @xmath14 . the cauchy problem for system ( [ driftcontrol ] )
is correctly stated for any locally integrable vector - function @xmath27 .
we assume , throughout this section , that system ( [ driftcontrol ] ) is complete , i.e. all solutions of the system are defined on the whole semi - axis @xmath220 .
this completeness assumption is automatically satisfied if one of the following is true : ( i ) if @xmath14 is a compact manifold , ( ii ) @xmath14 is a lie group and the fields @xmath221 are left - invariant , or ( iii ) if @xmath14 is a closed submanifold of the euclidean space and @xmath222 .
define the endpoint map @xmath223,{\mathbb{r}}^k)\to m$ ] by @xmath224 where @xmath45 is the admissible pair corresponding to the control system ( [ driftcontrol ] ) with initial condition @xmath166 .
it is known that the map @xmath225 is a smooth mapping .
the critical points of the map @xmath226 are called _
singular controls_. admissible paths corresponding to singular controls are called _
singular paths_. we also need the hessian of the mapping @xmath225 at the critical point .
( see @xcite for detail . )
let @xmath227 be a banach space which is an everywhere dense subspace of a hilbert space @xmath161 .
consider a mapping @xmath228 such that the restriction of this map @xmath229 to any finite dimensional subspace @xmath230 of the banach space @xmath227 is @xmath110 .
moreover , we assume that the first and second derivatives of all the restrictions @xmath229 are continuous in the hilbert space topology on the bounded subsets of @xmath227 . in other words ,
@xmath231 where @xmath232 is a linear map and @xmath233 is a quadratic mapping from @xmath227 to @xmath234 .
moreover , @xmath235 , @xmath236 and @xmath237 depend continuously on @xmath238 in the topology of @xmath161 while @xmath238 is contained in a ball of @xmath227 .
the hessian @xmath239 of the function @xmath240 is the restriction of @xmath233 to the kernel of @xmath232 with values considered up to the image of @xmath232 .
hessian is a part of @xmath233 which survives smooth changes of variables in @xmath227 and @xmath234 .
let @xmath241 be a covector in the dual space @xmath242 such that @xmath243 , then @xmath244 is a well - defined real quadratic form on @xmath245 .
we denote the morse index of this quadratic form by @xmath246 .
recall that the morse index of a quadratic form is the supremum of dimensions of the subspaces where the form is negative definite .
a critical point @xmath238 of @xmath240 is called _ sharp _ if there exists a covector @xmath247 such that @xmath243 and @xmath248 .
needless to say , the spaces @xmath249 and @xmath234 can be substituted by smooth manifolds ( banach , hilbert and @xmath65-dimensional ) in all this terminology .
going back to the control system ( [ driftcontrol ] ) , let @xmath45 be an admissible pair for this system .
we say that the control @xmath27 and the path @xmath46 are sharp if @xmath27 is a sharp critical point of the mapping @xmath250 .
one necessary condition for controls and paths to be sharp is the , so called , goh condition . [
goh](goh condition ) if @xmath251 , then @xmath252(x(t)))=0 \quad i , j=1,\ldots , k,\ 0\le t\le 1,\ ] ] where @xmath253 and @xmath254 is the local flow of the control system ( 20 ) with control equal to @xmath27 . see @xcite and references therein for the proof and other effective necessary and sufficient conditions of the sharpness .
now consider the optimal control problem @xmath255 where the infimum ranges over all admissible pairs @xmath40 corresponding to the control system ( [ driftcontrol ] ) with initial condition @xmath72 and final condition @xmath73 .
let @xmath55 be the hamiltonian function defined in ( [ maxham ] ) . for simplicity , we assume that the hamiltonian is @xmath110 .
a minimizer @xmath46 of the above problem is called normal if there exists a curve @xmath256\to t^*m$ ] in the cotangent bundle @xmath190 such that @xmath257 and @xmath258 is a trajectory of the hamiltonian vector field @xmath56 .
singular minimizers are also called abnormal . according to this not perfect terminology
a minimizer can simultaneously be normal and abnormal .
a minimizer which is not normal is called strictly abnormal .
the following theorem gives simple sufficient conditions for completeness of the cost function defined in ( [ costagain ] ) .
it is a combination of the well - known existence result ( see @xcite ) and necessary optimality conditions ( see @xcite ) .
( completeness of costs)[mintraj ] let @xmath20 be a lagrangian function which satisfies the following : 1 . @xmath20 is bounded below and there exist constants @xmath259 such that the ratio @xmath260 tends to 0 as @xmath261 uniformly on compact subsets of @xmath14 ; 2 . for any compact @xmath262 there exist constants @xmath263 such that @xmath264 , @xmath265 ; 3 .
the function @xmath266 is a strongly convex function for all @xmath267 .
then , for each pair of points @xmath136 in the manifold @xmath14 which satisfy @xmath268 , there exists an admissible pair @xmath45 achieving the infimum in ( [ costagain ] ) .
moreover , the minimizer @xmath46 is either a normal or a sharp path .
theorem [ mintraj ] gives lots of examples that satisfy condition ( 3 ) in theorem [ main ] .
in particular , this applies to the case where the control set @xmath269 and the lagrangian is @xmath270 .
next , we proceed to the main result of of this section which concerns with the lipschitz regularity of the cost function .
this takes care of the condition ( 2 ) in theorem [ main ] .
[ lip](lipschitz regularity ) assume that the system ( [ simplecontrol ] ) does not admit sharp controls and the lagrangian @xmath20 satisfies conditions of theorem [ mintraj ] , then the set @xmath271,{\mathbb{r}}^k)\}$ ] is open in the product @xmath115 .
moreover , the function @xmath272 is locally lipschitz on the set @xmath273 , where the cost @xmath12 is given by ( [ cost ] ) . in the case where the endpoint map is a submersion
, there is no singular control .
therefore , theorem [ lip ] is applicable .
in particular , this theorem , together with theorem [ main ] and [ mintraj ] , can be used to treat the cases considered in @xcite . in section 5
, we will consider a class of examples where the endpoint map is not necessarily a submersion , but theorem [ lip ] is still applicable .
the rest of the section is devoted to the proof of theorem [ lip ] . given @xmath238 in the banach space @xmath227
, we write @xmath274 if @xmath275 for any @xmath241 in @xmath276 such that @xmath243 .
it is easy to see that @xmath277 is an open subset of @xmath227 for any integer @xmath278 .
let @xmath279 be radius @xmath280 balls in @xmath227 and @xmath234 centered at @xmath238 and @xmath18 respectively .
the following is a qualitative version of openness of a mapping @xmath240 and any mapping @xmath281 close to it .
we say that the map @xmath228 is @xmath282-solid at the point @xmath238 of the banach space @xmath227 if for some constant @xmath283 and any sufficiently small @xmath284 , there exists @xmath285 such that @xmath286 for any @xmath287 such that @xmath288 .
implicit function theorem together with brouwer fixed point theorem imply that @xmath240 is 1-solid at any regular point . [ 2solid ] if @xmath289 then @xmath240 is 2-solid at @xmath238 .
this lemma is a refinement of theorem 20.3 from @xcite .
it can be proved by a slight modification of the proof of the cited theorem .
obviously , we may assume that @xmath238 is a critical point of @xmath240 .
moreover , by an argument in the proof of the above cited theorem , we may assume that @xmath227 is a finite dimensional space , @xmath290 and @xmath291 .
let @xmath292 , where @xmath293 .
for any @xmath294 we write @xmath295 , where @xmath296 .
now consider the mapping @xmath297 it is shown in the proof of theorem 20.3 from @xcite that @xmath298 contains regular points in any neighborhood of 0 . hence @xmath299 such that the image of any continuous mapping @xmath300 sufficiently close ( in @xmath281-norm ) to @xmath301 contains @xmath302 .
now we set @xmath303 ; then @xmath304 as @xmath305 and the desired property of @xmath240 is reduced to the already established property of @xmath306 .
the minimization problem ( [ costagain ] ) can be rephrased into a constrained minimization problem in an infinite - dimensional space .
for simplicity , consider the case where @xmath11 .
let @xmath45 be an admissible pair of the control system ( [ driftcontrol ] ) and let @xmath307,{\mathbb{r}}^k)\to{\mathbb{r}}$ ] be the function defined by @xmath308 let @xmath309,{\mathbb{r}}^k)\to { \mathbb{r}}^n\times { \mathbb{r}}^n$ ] be the map @xmath310 finding the minimum in ( [ costagain ] ) is now equivalent to minimizing the function @xmath198 on the set @xmath311 .
due to the above discussion , we consider the following general setting .
consider a function @xmath312 on the banach space @xmath227 such that @xmath313 is a @xmath110-mapping for any finite dimensional subspace @xmath230 of @xmath227 .
assume that the function @xmath198 as well as the first and second derivatives of the restrictions @xmath313 are continuous on the bounded subsets of @xmath227 in the topology of @xmath161 .
assume that @xmath148 is a bounded subset of @xmath227 that is compact in the topology of @xmath161 and satisfies the following property : @xmath314 for any @xmath238 in the set @xmath148 .
we define a function @xmath6 on @xmath315 by the formula @xmath316 .
[ lipschitz ] if @xmath317 for any @xmath318 , then @xmath6 is locally lipschitz . given @xmath238 in @xmath148 ,
there exists a finite dimensional subspace @xmath230 of the banach space @xmath227 such that @xmath319 . then @xmath320 .
hence @xmath321 is 2-solid at @xmath238 and @xmath322 for some c and any sufficiently small @xmath280 .
let @xmath323 and @xmath324 , then @xmath325 for some @xmath326 .
we have : @xmath327 moreover , the compactness of @xmath148 allows to chose unique @xmath328 and the bound for @xmath280 for all @xmath318 . in particular , we can exchange @xmath18 and @xmath19 in the last inequality .
hence @xmath329 we perform the proof only in the case @xmath330 in order to simplify the language .
generalization to any manifold is straightforward .
we set @xmath331,{\mathbb{r}}^k),\ h=\mathbb r^n\times l^2([0,t],{\mathbb{r}}^k),\ ] ] @xmath332 and apply the above results .
first of all , @xmath333 for all @xmath334 since our system does not admit sharp controls .
lemma [ 2solid ] implies that @xmath240 is 2-solid and @xmath335 is open .
now let @xmath336 be a ball in @xmath227 equipped with the weak topology of @xmath161 .
the _ endpoint mapping _
@xmath240 is continuous as a mapping from @xmath336 to @xmath337 .
strict convexity of @xmath20 implies that there is some constant @xmath283 such that @xmath338 as @xmath339 and @xmath340 .
therefore , @xmath341 and @xmath342 if and only if @xmath343 converges to @xmath334 in the strong topology of @xmath161 .
assume that @xmath344 for all @xmath65 .
inequality @xmath345 would imply that @xmath346 on the other hand , the openness of the map @xmath240 implies that @xmath347 hence @xmath342 and @xmath343 converges to @xmath334 in the strong topology of @xmath161 .
let @xmath348 be a compact subset of @xmath349 and @xmath350 then @xmath148 is contained in some ball @xmath336 .
recall that @xmath336 is equipped with the weak topology ; it is compact
. now calculations of previous 2 paragraphs imply compactness of @xmath148 in the strong topology of @xmath161 .
finally , we derive the lipschitz property of @xmath351 from lemma [ lipschitz ] .
in this section , we will apply the results in the previous sections to some subriemannian manifolds .
first , let us recall some basic definitions .
let @xmath68 and @xmath352 be two ( possibly singular ) distributions on a manifold @xmath14 . define the distribution @xmath353 $ ] by @xmath354_x=\text{span}\{[v , w](x)|v \text { is a section of } \delta , w \text { is a section of } \delta'\}.\ ] ] define inductively the following distributions : @xmath355=\delta^2 $ ] and @xmath356 $ ] .
a distribution @xmath68 is called @xmath76-generating if @xmath357 and the smallest such @xmath76 is called the degree of nonholonomy . also , the distribution is called bracket generating if it is @xmath76-generating for some @xmath76 .
if @xmath68 is a bracket generating distribution , then it defines a flag of distribution by @xmath358 the growth vector of the distribution @xmath68 at the point @xmath18 is defined by @xmath359 .
the distribution @xmath68 is called regular if the growth vector is the same for all @xmath18 .
let @xmath360\to m$ ] be an admissible curve , that is a lipschitz curve almost everywhere tangent to @xmath68 .
the following classical result on bracket generating distributions is the starting point of subriemannian geometry .
( chow and rashevskii ) given any two points @xmath18 and @xmath19 on the manifold @xmath14 with a bracket generating distribution , there exists an admissible curve joining the two points . using chow - rashevskii theorem
, we can define the subriemannian distance @xmath1 .
let @xmath91 be a fibre inner product on the distribution @xmath68 , called subriemannian metric .
the length of an admissible curve @xmath46 is defined in the usual way : @xmath361 .
the subriemannian distance @xmath362 between two points @xmath18 and @xmath19 is defined by the infimum of the length of all admissible curves joining @xmath18 and @xmath19 .
there is a quantitative version of chow - rashevskii theorem , called ball - box theorem , which gives hlder continuity of the subriemannian distance .
see @xcite for detail .
[ lip2 ] let @xmath95 be the metric of a complete subriemannian space with distribution @xmath68 .
function @xmath363 is locally lipschitz if and only if the distribution is 2-generating .
the systems with 2-generating distributions do not admit sharp paths because these systems are not compatible with the goh condition . on the other hand , constant paths ( points ) are sharp minimizers in the case of distributions whose nonholonomy degree is greater than 2 and the ball - box theorem implies that @xmath0 is not locally lipschitz at the diagonal in this case .
combining corollary [ lip2 ] with theorem [ main ] , we prove the existence and uniqueness of optimal map for subriemannian manifold with 2-generating distribution .
[ maintransport ] let @xmath14 be a complete subriemannian manifold defined by a 2-generating distribution , then there exists a unique ( up to @xmath6-measure zero ) optimal map to the monge s problem with the cost @xmath12 given by @xmath96 . here
@xmath95 is the subriemannian distance of @xmath14 . the locally lipschitz property of the distance @xmath1 out of the diagonal is guaranteed for much bigger class of distribution . in particular , it is proved in @xcite that generic distribution of rank @xmath364 does not admit non - constant sharp trajectories . in the class of carnot groups ,
the following estimates are valid : generic @xmath65-dimensional carnot group with rank @xmath76 distribution does not admit nonconstant sharp trajectories if @xmath365 and has nonconstant sharp length minimizing trajectories if @xmath366 . recall that a simply - connected lie group endowed with a left - invariant distribution @xmath367 is a carnot group if the lie algebra @xmath368 is a graded nilpotent lie algebra such that it is lie generated by the block with lowest grading ( i.e. @xmath369 , @xmath370=v_{i+j}$ ] , @xmath371 and @xmath367 lie - generates @xmath368 ) . clearly , if the cost is locally lipschitz out of the diagonal , then the statement of theorem 4.1 remains valid with the extra assumption that the supports of the initial @xmath6 and the final measures @xmath7 are disjoint : @xmath372 .
according to theorem [ lip ] , it remains to study the case where sharp controls exist . in this section , we will prove a property of optimal map when the cost is continuous .
normal minimizers will play a very important role .
we continue to study optimal control problem ( 20 ) , ( 21 ) .
as we already mentioned , strictly abnormal minimizers must be sharp .
in addition , if @xmath373 , then optimal control cost is continuous . according to the discussion at the end of the previous section ,
we expect strictly abnormal minimizers mainly for generic rank 2 distributions on the manifold of dimension greater than 3 and for generic carnot groups of big enough corank . in these situations ,
strictly abnormal minimizers are indeed unavoidable .
the existence of strictly abnormal minimizers for subriemannian manifolds is first done in @xcite . in @xcite and @xcite
, it is shown that there are many strictly abnormal minimizers in general for subriemannian manifolds .
( see , for instance , theorem [ liusus ] below . ) finally , a general theory on abnormal minimizers for rank 2 distributions is developed in @xcite .
see @xcite for a detail account on the history and references on abnormal minimizers . here is a sample result in @xcite which is of interest to us .
[ liusus](liu and sussman ) let @xmath14 be a 4-dimensional manifold with a rank 2 regular bracket generating distribution @xmath68 and subriemannian metric @xmath91 .
let @xmath374 and @xmath375 be two global sections of @xmath68 such that 1 .
@xmath374 and @xmath375 are everywhere orthonormal , 2 .
@xmath374 , @xmath375 , @xmath376 $ ] and @xmath377 $ ] are everywhere linearly dependent , 3 .
@xmath375 , @xmath376 $ ] and @xmath377 $ ] are everywhere linearly independent .
then any short enough segments of the integral curves of the vector field @xmath375 are strictly abnormal minimizers .
we call a local flow a strictly abnormal flow if the corresponding trajectories are all strictly abnormal minimizers
. an interesting question is whether time-1 map of an abnormal flow is an optimal map .
the following theorem shows that this is not the case for any reasonable initial measure and continuous cost .
[ normopt ] assume that the cost @xmath12 in ( [ optimal ] ) is continuous , bounded below and the support of the measure @xmath6 is equal to the closure of its interior . if @xmath378 is a continuous map such that @xmath379 achieves the infimum in problem [ kanre ] , then @xmath18 and @xmath380 are connected by a normal minimizer on a dense set of @xmath18 in the support of @xmath6 .
by theorem [ dualexist ] , there exists a function @xmath381 such that @xmath58 and its @xmath120-transform achieve the supremum in problem [ dual ] .
moreover , by theorem [ dualregular ] , the functions @xmath58 and @xmath121 are upper semicontinuous . by theorem [ existjoint ] , @xmath382 for @xmath6-almost all @xmath18 . by upper semicontinuity of @xmath58 and @xmath121 , @xmath383 but @xmath384 for any @xmath385 in the manifold @xmath14
so , ( [ normopt1 ] ) holds for all @xmath18 in the support @xmath24 of @xmath6 .
therefore , @xmath18 achieves the infimum @xmath386 $ ] for all @xmath18 in the support of @xmath6 .
moreover , using ( [ normopt1 ] ) , it is easy to see that the function @xmath58 is continuous on @xmath24 . in particular , it is subdifferentiable on a dense set of @xmath24 . by proposition [ dualmin ] and theorem [ apppmp ] , @xmath18 and @xmath380
is connected by a normal minimizer if @xmath58 is subdifferentiable at @xmath18 .
this proves the theorem .
in this section , we describe an important class of control systems which admit smooth optimal maps built essentially from abnormal minimizers .
recall that abnormal minimizers are singular trajectories of the control system whose definition does not depend on the lagrangian .
let @xmath387 be a smooth principal bundle where the structural group @xmath388 is a connected abelian lie group .
let @xmath389 be the vertical vector fields which generate the action of @xmath388 .
consider the following control system @xmath390 where @xmath391 is a smooth vector field on @xmath14 , and the re - scaled systems @xmath392 for @xmath284 .
we define the hamiltonian @xmath393 by @xmath394 where @xmath395 is a covector in @xmath396 .
we assume that the maximum above is achieved for any @xmath241 in @xmath396 and is finite .
typical example is the hopf bundle @xmath397 and a left - invariant vector field @xmath398 .
then @xmath399 , where @xmath60 is a constant and @xmath400 is the length of the covector @xmath241 with respect to the standard ( constant curvature ) riemannian structure on the sphere .
( see ( * ? ? ? *
section 22.2 ) ) consider the following control system on @xmath401 with admissible pair @xmath402 contained in the @xmath388-bundle @xmath387 and admissible trajectory @xmath403 ( see remark [ generalcontrol ] ) : @xmath404 the function @xmath161 in ( [ abham ] ) is the hamiltonian of the time - optimal problem of the control system ( [ reducedcontrol ] ) .
( recall that the time optimal problem is the following minimization problem : fix two points @xmath42 and @xmath43 in @xmath401 and minimize the time @xmath405 among all admissible trajectories @xmath46 of the control system ( [ reducedcontrol ] ) such that @xmath406 and @xmath407 . )
system ( [ reducedcontrol ] ) is the reduced system associated to system ( [ controlagainagain ] ) according to the reduction procedure described in ( * ? ? ?
* chapter 22 ) .
in particular , @xmath408 transforms any admissible trajectory of system ( [ controlagainagain ] ) to the admissible trajectory of system ( [ reducedcontrol ] ) .
also , the smooth extremal trajectories of the time - optimal problem for system ( [ reducedcontrol ] ) are images under the map @xmath408 of singular trajectories of system ( [ controlagainagain ] ) . for any @xmath284 and any @xmath110 smooth function @xmath409
, we introduce the map @xmath410 where @xmath411 is the standard projection and @xmath412 is the hamiltonian flow of @xmath161 . set @xmath413 assume that @xmath414 pushes the measure @xmath415 forward to another measure @xmath416 on @xmath401 .
consider some `` lifts '' @xmath6 and @xmath7 of the measures @xmath415 and @xmath416 : @xmath417 .
let @xmath418 be an optimal map pushing forward @xmath6 to @xmath7 , then the following theorem says that @xmath419 is a covering of @xmath420 : @xmath421 by the discussion above , we see that @xmath18 and @xmath422 are connected by singular trajectories as claimed .
let @xmath148 be a compact subset of @xmath401 and @xmath423 .
assume that @xmath424 .
let @xmath6 and @xmath7 be borel probability measures such that @xmath425 .
then , for any sufficiently small @xmath284 and any optimal borel map @xmath426 of the control system ( [ controlrescaled ] ) with any lagrangian @xmath20 , the following is true whenever @xmath427 : @xmath428 we start from the following .
we say that a borel map @xmath429 is @xmath280-admissible for system ( [ controlrescaled ] ) if there exists a borel map @xmath430,g)$ ] such that @xmath431 where @xmath432 is an admissible trajectory of the reduced control system ( [ reducedcontrol ] ) with initial condition @xmath433 .
we are going to prove that @xmath420 is an admissible map , unique up to a @xmath434-measure zero set , which transforms @xmath434 into @xmath435 .
this fact implies the statement of the theorem .
inequality @xmath436 implies that @xmath437 is transversal to the level hypersurface of @xmath438 through @xmath18 .
hence the map @xmath420 is invertible on a neighborhood of @xmath148 for any sufficiently small @xmath280 .
moreover , the curve @xmath439 , is a unique admissible trajectory of system [ reducedcontrol ] which starts at the hypersurface @xmath440 and arrives at the point @xmath441 at time moment not greater than @xmath280 .
the last fact is proved by a simple adaptation of the standard sufficient optimality condition ( see ( * ? ? ?
* chapter 17 ) ) .
now we set @xmath442 then @xmath443 is a smooth function defined on a neighborhood of @xmath148 .
optimality property of @xmath420 implies that @xmath444 for any @xmath280-admissible map @xmath306 and any @xmath445 , and the inequality is strict at any point @xmath18 where @xmath446 .
in particular , if @xmath447 then @xmath448 @xmath449 hence @xmath450
grushin plane is the subriemannian space with base space @xmath451 and a singular distribution defined by the span of the following vectors @xmath452 in each tangent space . in other word
, the fibre of this distribution is the whole tangent space of @xmath451 if @xmath453 and it is spanned by @xmath454 otherwise .
we define a subriemannian metric by declaring that the two vector fields above are orthonormal .
the control system is given by @xmath455 the subriemannian distance @xmath1 is given by @xmath456 . in this section ,
we consider the optimal transport problem with cost @xmath12 given by @xmath78 .
there is no abnormal minimizer for this problem , so we consider its hamiltonian @xmath161 given by @xmath457 the corresponding hamiltonian equation is @xmath458 for simplicity , we consider the case @xmath459 . and we let @xmath460 and @xmath461 . in this case , the solutions give geodesics emanating from a point @xmath462 on the @xmath19-axis .
they are parameterized by @xmath463 and are given by @xmath464 if @xmath465 and given by @xmath466 if @xmath467 .
a geodesic is length minimizing if and only if @xmath468 .
= 10 cm next , we consider the mass transport problem .
let @xmath1 be the subriemannian distance of the grushin plane and consider problem [ optimal]with cost @xmath12 given by square of the subriemannian distance @xmath0 .
we also specialize to the case where the target measure @xmath7 is equal to the delta mass supported at the origin . in this case , the optimal map is clearly given by the constant map @xmath469 .
we are interested in the displacement interpolation corresponding to this optimal map .
recall that displacement interpolation is the one parameter family of maps @xmath470 such that @xmath470 is the optimal map with the cost @xmath471 given by the following : @xmath472 where the infimum ranges over all admissible pairs @xmath45 of the control system ( [ control ] ) with initial condition @xmath72 and final condition @xmath473 .
it is easy to see that if @xmath474 as in theorem [ main ] , then the displacement interpolation @xmath470 is given by @xmath475 .
moreover , the displacement interpolation is related to the hamilton - jacobi equation via the method of characteristics .
see @xcite and @xcite for details . to do this ,
we first evaluate the equations ( [ geo1 ] ) and ( [ geo2 ] ) at @xmath476
. then we solve @xmath438 and @xmath477 in terms of @xmath478 and @xmath479 . if @xmath480 is the function defined by @xmath481 , then @xmath58 is invertible .
a computation shows that @xmath482 therefore , the displacement interpolation is given by @xmath483 where @xmath484 and @xmath485 are given by @xmath486
this appendix is devoted to the prove of theorem [ pmpb ] .
the first step is to reduce the problem into a simpler one . recall that the bolza problem is the following minimization problem : @xmath51 where the infimum is taken over all admissible pair @xmath45 satisfying the control system @xmath487 and initial condition @xmath166 .
let @xmath488 be a point in the product manifold @xmath489 and consider the following extended control system on it : @xmath490 note that @xmath491 satisfies this extended system and initial condition @xmath492 if and only if @xmath46 satisfies the original control system in the bolza problem with the initial condition @xmath166 and @xmath493 .
therefore , problem [ bolza ] is equivalent to the following problem .
[ extendmin ] @xmath494 where the infimum is taken over all admissible pair satisfying the extended control system ( [ extendcontrol ] ) .
problem [ extendmin ] is an example of the mayer problem .
let @xmath495 be a function on the manifold @xmath401 and the mayer problem is the following minimization problem : [ mayer ] @xmath496 where the infimum is taken over all admissible pair @xmath497 satisfying the control system @xmath498 on @xmath401 and initial condition @xmath499 . for each point @xmath35 in the control set @xmath24 , define the corresponding hamiltonian function @xmath500 by @xmath501 [ pmp](pontryagin maximum principle for mayer problem ) let @xmath502 be an admissible pair which achieve the infimum in problem [ mayer ] .
assume that the function @xmath503 in problem [ mayer ] is super - differentiable at the point @xmath504 and let @xmath505 be in the super - differential @xmath506 of @xmath503 .
then there exists a lipschitz path @xmath507\to t^*n$ ] which satisfies the following for almost all time @xmath41 in the interval @xmath25 $ ] : @xmath508 fix a point @xmath238 in the control set and a number @xmath509 in the interval @xmath25 $ ] . for each small positive number @xmath142 ,
let @xmath510 be the admissible control defined by @xmath511 $ ; } \\
v , & \hbox{if $ t\in[\tau-{\epsilon},\tau]$. } \\
\end{array } \right.\ ] ] since the optimal control @xmath512 is locally bounded , the new control @xmath510 defined above is also locally bounded .
let @xmath513 be the time - dependent local flow of the following ordinary differential equation @xmath514 here , @xmath515 denotes the image of the point @xmath516 in the manifold @xmath401 under the local flow @xmath517 at time @xmath41 .
it has the property that @xmath518 . also , recall that @xmath519 depends smoothly on the space variable , lipschitz with respect to the time variables .
since @xmath520 and the function @xmath503 is minimizing at @xmath504 , the following is true for all @xmath142 : @xmath521 let @xmath505 be a point in the super - differentiable @xmath506 at the point @xmath504 , then there exists a @xmath522 function @xmath523 such that @xmath524 and @xmath525 has a local maximum at @xmath504 . combining this with ( [ pmp1 ] ) , we have @xmath526 @xmath527 simplifying this equation , we get @xmath528 if @xmath529 denotes the flow of the vector field @xmath530 , then @xmath531 so , if we assume that @xmath509 is a point of differentiability of the map @xmath532 which is true for almost all time @xmath509 in the interval @xmath25 $ ] , then @xmath533 is differentiable with respect to @xmath534 at zero . therefore , we can let @xmath534 goes to 0 in ( [ pmp2 ] ) and obtain @xmath535 if we differentiate equation ( [ pmp3 ] ) with respect to @xmath534 and set it to zero , it becomes @xmath536 substitute this equation back into ( [ pmp4 ] ) , we get the following : @xmath537 define @xmath62\to t^*n$ ] by @xmath538 , then the first two assertions of the theorem are clearly satisfied . the following is well known ( see @xcite or @xcite ) . [ pretau ]
let @xmath539 be the tautological 1-form on the cotangent bundle of the manifold @xmath401 , then for each diffeomorphism @xmath540 , the pull back map @xmath541 on the cotangent bundle of the manifold preserves the 1-form @xmath542 .
let @xmath543 be the time - dependent vector field on the cotangent bundle of the manifold which satisfies @xmath544 for almost all time @xmath41 in @xmath25 $ ] . if @xmath545 denotes the lie derivative with respect to a vector field @xmath80 , then , by lemma [ pretau ] , the following is true for almost all time @xmath41 in @xmath25 $ ] : @xmath546 if @xmath547 is the canonical symplectic 2-form on the cotangent bundle , then , by using cartan s formula , we have @xmath548 therefore , the vector field @xmath543 is a hamiltonian vector field with hamiltonian given by @xmath549 the third assertion of the theorem follows from this .
the last assertion follows from ( [ pmp5 ] ) .
going back to problem [ extendmin ] , we can apply pontryagin maximum principle for mayer problem .
let @xmath550 be an admissible pair which minimizes problem [ extendmin ] and let @xmath551 be the function defined by @xmath552 by theorem [ pmp ] , there exists a curve @xmath553\to t^*_{{\widetilde{x } } } m\times{\mathbb{r}}$ ] such that @xmath554 and @xmath555 from the first equation in ( [ extendham ] ) , we get @xmath556 .
so , @xmath557 . therefore , ( [ extendham ] ) is simplified to @xmath558 this finishes the proof of theorem [ pmpb ] .
the second author would like to express deep gratitude to his supervisor , boris khesin , who suggested to him the problem of optimal mass transportation on subriemannian manifolds .
100 l. ambrosio , s. rigot : optimal mass transportation in the heisenberg group , j. func .
208(2004 ) , 261 - 301 a. a. agrachev : geometry of optimal control problems and hamiltonian systems , lecture noes , 2004 a. a. agrachev , j.p .
gauthier , on the subanalyticity of carnot - caratheodory distances , ann .
i. h. poincar an 18 , ( 2001 ) , 359382 a. a. agrachev , y. l. sachkov : control theory from the geometric viewpoint , springer , 2004 a. a. agrachev , a. v. sarychev : strong minimality of abnormal geodesics for 2-distributions , jdcs , 1995 , v.1 , 139 - 176 a. a. agrachev , a. v. sarychev : abnormal sub - riemannian geodesics , morse index and rigidity , annales de linstitut henry poincar - analyse non linaire , v.13 , 1996 , 635 - 690 p. bernard , b. buffoni : optimal mass transportation and mather theory , 2004 , preprint y. brenier : polar factorization and monotome rearrangement of vecotr - valued functions , comm .
pure appl .
44 , 4(1991 ) , 323 - 351 p. cannarrsa , l. rifford : semiconcavity results for optimal control problems admitting no singular minimizing controls , ann .
inst . h. poincare anal .
non lineaire p. cannarrsa , c. sinestrari : semiconcave functions , hamilton - jacobi equations , and optimal control , birkhaser boston , 2004 a. fathi , a. figalli : optimal transportation on non - compact manifolds , preprint a. figalli : existence , uniqueness and regularity of optimal transport maps , preprint r. v. gamkrelidze : principles of optimal control theory .
plenum publishing corporation , new york , 1978 l. kantorovich : on the translocation of masses , c.r .
( doklady ) acad .
urss(n.s . )
, 37 , 1942 , 199 - 201 w.s .
liu , h.j .
sussman : shortest paths for sub - riemannian metrics on rank-2 distributions , memoirs of ams , v.118 , n. 569 , 1995 j. marsden , t. ratiu : introduction to mechanics and symmetry , springer , 1999 r. mccann : polar factorization of maps in riemannian manifolds , geometric and functional analysis , vol.11 , 2001 r. montgomery : abnormal minimizers , siam j. control and optimization , vol .
32 , no . 6 , 1994 , 1605 - 1620 .
r. montgomery : a tour of subriemannian geometries , their geodesics and applications , ams , 2002 a. sarychev , d. torres : lipschitzian regularity conditions for the mnimizing trajectories of optimal control problems , nonlinear analysis and its applications to differential equations ( lisbon , 1998 ) , 357 - 368 , progr .
nonlinear differential equations appl .
, 43 , birkhaser boston , boston , ma , 2001 h.j .
sussman : a cornucopia of abnormal sub - riemannian minimizers .
part i : the four dimensional case , i m a technical report no .
1073 , december , 1992 c. villani : topics in mass transportation , ams , providence , rhode island , 2003 c. villani : optimal transport : old and new , preprint | we study monge s optimal transportation problem , where the cost is given by optimal control cost .
we prove the existence and uniqueness of an optimal map under certain regularity conditions on the lagrangian , absolute continuity of the measures with respect to lebesgue , and most importantly the absence of sharp abnormal minimizers .
in particular , this result is applicable in the case of subriemannian manifolds with a 2-generating distribution and cost given by @xmath0 , where @xmath1 is the subriemannian distance .
also , we discuss some properties of the optimal plan when abnormal minimizers are present . finally , we consider some examples of displacement interpolation in the case of grushin plane . |
quantum technology promises to provide a robust platform combining otherwise contrasting degrees of freedom to achieve various tasks of quantum information and computation .
light - matter coupled quantum systems have received particular attention in the bid to engineer scalable quantum platforms @xcite . in these hybrid systems a two - level system ( qubit )
is coupled to a bosonic field . theoretically , the basic building block for describing such light - matter systems is the quantum rabi model , which takes the following form under the dipole approximation @xcite , @xmath0 here , @xmath1 and @xmath2 are the annihilation and creation operators for the bosonic field of frequency @xmath3 , @xmath4 with @xmath5 the pauli matrices for the two - level system , @xmath6 is the energy level splitting between the two levels , and @xmath7 denotes the coupling strength between the bosonic mode and the two - level system ( assumed to be positive ) . in conventional cavity qed settings
the light - matter coupling strength @xmath7 is several orders of magnitude smaller than the transition frequencies @xmath8 .
in addition , if the system is near resonance , such that @xmath9 , the full rabi model can be simplified by applying the rotating wave approximation ( rwa ) . under this approximation the so - called `` counter - rotating '' terms in eq . can be neglected , leading to the much simpler and readily solvable jaynes - cummings ( jc ) model @xcite @xmath10 however , the validity of the rabi model is not only restricted to cavity qed setups . the original rabi model is a ubiquitous physical model capable of describing a wide variety of other physical systems , including trapped ions @xcite , qubit - coupled nanomechanical resonators @xcite , and circuit qed systems @xcite .
circuit qed architectures are particularly interesting because it is possible to reach physical regimes where the light - matter coupling strength @xmath7 becomes a sizeable fraction of the transition frequency of the boson field and/or the two - level system , meaning that the rwa is no longer a valid approximation .
the influence of non - rwa terms has been studied theoretically for some time @xcite . however , with recent experimental advances in reaching `` ultra - strong '' and `` deep strong '' coupling regimes of light and matter @xcite , investigations of the full rabi model are attracting increasing attention @xcite .
it has been clearly shown that in these strongly coupled quantum systems the simplified jc model no longer applies and it becomes necessary to consider the full rabi model to capture the relevant physics .
the significance of non - rwa terms has also been elucidated in many - body extensions of the full rabi model , both in the equilibrium @xcite and non - equilibrium @xcite settings .
another regime of significant practical interest in which the rabi model can be substantially simplified is the so - called `` dispersive regime '' @xcite . in the dispersive limit , the qubit and the boson field are far detuned compared to the light - matter coupling strength @xmath7 ,
_ i.e. _ @xmath11 .
this regime is widely considered in experiments , particularly in circuit qed , as it allows a non - demolition type measurement of the qubit by probing the resonator @xcite .
although the dispersive approximation is often applied together with the rwa @xcite , in this work we consider the dispersive limit without making the rwa @xcite and show that it can be used to create squeezed states of the field mode .
the generation of squeezed light has attracted much attention for various applications , including high - precision quantum measurements @xcite and quantum communication @xcite . in the quantum optical domain ,
squeezed light has been more commonly generated using nonlinear optical processes , including degenerate parametric amplification and degenerate four - wave mixing @xcite .
these nonlinear processes require large optical nonlinearities , low intracavity and detection losses , and low phase noise @xcite . over the years
, experimental progress has made it possible to generate squeezed states of light in a variety of different experimental setups @xcite , including optical parametric oscillators @xcite , superconducting cavities @xcite , and optomechanical cavities @xcite . in this work
we embark on an alternative route to generate squeezed states of a cavity field mode dispersively coupled to a two - level system .
it has been shown that in the dispersive regime beyond the rwa , the ground state of the rabi model exhibits one - mode squeezing of the boson field @xcite , but the degree of squeezing is very small . however
, it has been known for a long time that any nonadiabatic change in the frequency of a harmonic oscillator ( boson field ) results in squeezing of the state of the oscillator @xcite .
the degree of squeezing is particularly pronounced if the frequency change is sudden @xcite .
moreover , it is possible to use periodic sudden jumps between two frequencies to produce arbitrarily large squeezing of the field mode @xcite .
we show that this strategy can be used in the context of the dispersive quantum rabi model to generate significant squeezing of the field mode . comparing analytical predictions using the dispersive approximation with numerical simulations of the same protocol using the full rabi hamiltonian
, we find that , remarkably , an even larger degree of squeezing arises in the latter case .
an analysis of the effects of noise and imperfections suggests that considerable squeezing could be achieved with existing experimental capabilities in circuit qed .
the paper is organized as follows . in sec.[sec : disp ] we introduce the dispersive theory describing the interaction between a quantized bosonic mode and a single two - level system beyond the rwa .
we evaluate the degree of squeezing of the field mode present in the ground state of the dispersive hamiltonian and compare it with that of the ground state of the full rabi model . in sec .
[ sec : grndstsqzngprot ] we present a protocol that can significantly enhance the degree of squeezing of the field mode .
the influence of noise and imperfections is examined in sec .
[ sec : noiseevl ] and the paper concludes with a short discussion in sec . [ sec : outlk ] .
the physical setting we consider is a strongly coupled light - matter system modeled by the rabi hamiltonian of eq . .
throughout this paper we focus on the dispersive regime in which the detuning @xmath12 between the qubit and the cavity is large compared to their coupling , @xmath13 .
furthermore , we will assume that @xmath14 and therefore will not invoke the rwa to simplify the light - matter interaction .
although the theory presented here may describe many different types of experimental systems , the dispersive limit is particularly applicable to circuit qed experiments @xcite , which we will touch on near the end of the paper .
an effective hamiltonian in the dispersive limit may be derived @xcite using the unitary transformation @xmath15 , where @xmath16 , @xmath17 , and @xmath18 denotes the hermitian conjugate . applying this transformation to the rabi hamiltonian and keeping terms up to second order in @xmath7 , the dispersive hamiltonian @xmath19 is given by @xmath20 the eigenspectra of the dispersive hamiltonian and the full rabi model
are compared in fig .
[ eigvales ] .
the lowest lying energy levels of each model are plotted as a function of the detuning parameter @xmath21 for two different values of the light - matter coupling strength @xmath7 . as is clear from fig .
[ eigvales ] , the dispersive theory is a valid approximation to the full rabi model in the large detuning regime but breaks down as expected near @xmath22 . the mismatch between the two becomes more pronounced for larger values of @xmath7 as @xmath23 .
( dotted ) and the analytical solutions of the dispersive hamiltonian [ eq . ] ( solid ) for ( a ) @xmath24 and ( b ) @xmath25 .
note that @xmath26.,scaledwidth=49.0% ] an alternative form of eq . makes the physics of the dispersive interaction more transparent . defining the parameter @xmath27
, the dispersive hamiltonian may be re - expressed as @xmath28 in this form it is evident that the first term contains a shift in the frequency of the cavity mode that depends on the state of the qubit through @xmath29 .
this is the basis for the commonly used dispersive readout technique for superconducting qubits @xcite : the frequency shift of the cavity , which is readily measured , is correlated with the qubit state . as pointed out in @xcite , a remarkable feature of the above hamiltonian is that a dispersive readout of the cavity field is still possible even when the cavity field and the two - level system are coupled strongly enough that the rwa can not be made .
the final term of the dispersive hamiltonian is also of interest .
it takes the form of a one - mode squeezing interaction , the sign of which again depends on the state of the qubit through @xmath29 @xcite .
this suggests that the dispersive hamiltonian may be used to generate non - classical states of the field mode , which is the main focus of the present work . to begin with ,
we consider the degree of squeezing intrinsic to the ground state of the dispersive hamiltonian and , from there , the corresponding approximate ground state of the rabi hamiltonian @xcite . noting that @xmath29 is a constant of motion of the dispersive hamiltonian , the ground state of eq .
can be readily obtained by diagonalizing it in the subspace of the qubit states @xmath30 which denote , respectively , the @xmath31 and @xmath32 eigenstates of @xmath33 : @xmath34 where @xmath35 to diagonalize the hamiltonians @xmath36 we define a unitary transformation @xmath37 where @xmath38 is a unitary squeezing operator @xcite . under the action of this unitary transformation the annihilation and creation operators @xmath39 transform as @xmath40 where the squeezing parameter is defined as @xmath41 after the transformation @xmath42 is applied , the hamiltonian of eq . takes the diagonal form @xmath43 where @xmath44 the resulting hamiltonian is that of a harmonic oscillator whose frequency shift is correlated with the qubit state @xmath45 or @xmath46 . clearly , in order for the cavity mode to maintain its harmonic behavior it is necessary to have @xmath47 or , equivalently , @xmath48 .
this places a further restriction on the magnitude of the coupling strength @xmath7 , which is in addition to the requirement that @xmath49 in order for the truncation at @xmath50 used in deriving eq . to be valid .
the eigenstates of @xmath51 are simply given by @xmath52 and @xmath53 .
the dispersive hamiltonian consequently has the corresponding eigenstates @xmath54 hence the ground states of the dispersive hamiltonians @xmath36 take a separable form with the cavity field in a squeezed vacuum state .
since a unitary transformation leaves eigenvalues unchanged , the eigenvalues of the dispersive hamiltonian may be taken as a direct approximation to the eigenvalues of the full rabi model .
however , the same is _ not _ true of the eigenstates @xcite .
the eigenstates of the original rabi model are related to the eigenstates of the dispersive hamiltonian through the unitary transformations @xmath42 and @xmath55 : @xmath56 to maintain consistency with the dispersive hamiltonian , which is valid to second order in @xmath7 , we expand the operator @xmath55 to first order in @xmath57 ( which are proportional to @xmath7 ) : @xmath58 for @xmath59 , the corresponding approximate eigenstates of the rabi model are @xmath60 for @xmath61 the state @xmath62 is an approximation to the ground state of the rabi hamiltonian ; we have numerically confirmed this for the parameter values used throughout the work .
the state @xmath63 is a little more intriguing as it corresponds to one of the higher excited states of the rabi model ; which excited state depends on the value of @xmath21 .
we have numerically confirmed that for values of @xmath64 , @xmath63 is an approximation to the @xmath65 excited state when @xmath66 ( @xmath67 ) , @xmath68 ( @xmath69 ) and @xmath70 ( @xmath71 ) .
these states are not separable as the dispersive eigenstates are ; rather , they represent entangled states of the qubit and the cavity field . what is more , only the first term in each superposition , albeit the dominant one , contains a squeezed vacuum in the field .
the second term is a squeezed number or fock state , which is an even more highly nonclassical state than the squeezed vacuum .
nevertheless , a squeezed vacuum state of the field may be recovered by making a projective measurement onto the appropriate qubit state .
the differences between the dispersive eigenstates and the approximate rabi eigenstates have important consequences for the generation of squeezed states of the field , as we shall see later on . in order to quantify the degree of squeezing of the cavity field present in the the dispersive and rabi ground states , we introduce dimensionless position and momentum quadratures for the mode @xmath1
: @xmath72 first we consider the ground state in each qubit subspace of the dispersive hamiltonian , @xmath73 , for which the variances in the position and momentum quadratures are @xmath74 alternatively , in terms of the squeezing parameter given by eq .
the variances may be written as @xmath75 it is clear that , for @xmath76 , i.e. @xmath61 , the squeezing parameter @xmath77 ( which corresponds to the ground state of eq . ) results in noise reduction beyond the standard quantum limit in the momentum quadrature and enhanced fluctuations in the position quadrature .
the reverse is true for squeezing parameter @xmath78 .
hence the quadrature of squeezing depends on the sign of @xmath79 . in experiments
the degree of squeezing @xmath80 is commonly expressed in decibels ( db ) , calculated as @xmath81 .
the degree of squeezing present in the ground states of both qubit subspaces of the dispersive hamiltonian @xmath82 and in the approximate ground state of the rabi hamiltonian @xmath83 is plotted in fig .
[ sqznggrnd ] as a function of the light - matter coupling @xmath7 for three different values of the detuning @xmath21 . here and in what follows we concentrate on the case @xmath61 , which gives both a larger degree of squeezing and a larger frequency shift of the cavity field for a given value of @xmath84 .
figure [ sqznggrnd ] shows that the degree of squeezing in the ground state of the dispersive hamiltonian increases with the coupling @xmath7 .
this is expected since the parameter @xmath85 which controls the degree of squeezing scales as @xmath86 .
similarly , the squeezing is reduced as the detuning @xmath21 increases .
hence there is a tradeoff between the validity of the dispersive approximation , which requires @xmath87 , and the amount of squeezing that is present in the ground state . in any event ,
the degree of squeezing of the ground state of the dispersive hamiltonian is not large , @xmath88 db .
it can be easily verified that the squeezing parameter @xmath89 , which is why the degree of squeezing of @xmath90 is marginally higher than the degree of squeezing of @xmath91 . as a function of the light - matter coupling strength @xmath7 for the dispersive eigenstates @xmath92 ( thin solid , blue)and @xmath93 ( thick solid , red ) , the approximate ground state of the rabi model @xmath94 ( crosses , pink ) , and the numerically determined ground state of the rabi model @xmath95 ( filled - dots , black ) for @xmath96 , and @xmath97.,scaledwidth=49.0% ] for the sake of comparison the degree of squeezing in the exact ground state of the full rabi model , calculated numerically , is also shown in fig . [ sqznggrnd ] . from the figure it is evident that as the parameter @xmath85 decreases both the dispersive ground state @xmath98 and the approximate rabi ground state @xmath62 better approximate the exact ground state of the rabi model in terms of capturing the degree of squeezing @xmath80 .
the degree of squeezing in the approximate and the exact ground states of the rabi model is smaller than the degree of squeezing present in the dispersive states @xmath99 .
this can be understood by examining the form of eq .
, which is a superposition of a squeezed vacuum state @xmath100 and a squeezed @xmath101 fock state @xmath102 . for @xmath103 , the fock states @xmath104 are not states of minimum uncertainty ;
their variances are given by @xmath105 . therefore the state @xmath106 has variances @xmath107 and @xmath108 .
since the @xmath109 and @xmath101 states are associated with orthogonal qubit components , there is no coherence between them and the overall variances are equal to the sums of the variances of the two squeezed fock states . as a consequence , both quadrature variances are increased over those of the squeezed vacuum and hence the degree of squeezing of the overall state is reduced .
in the previous section we evaluated the degree of squeezing in the respective ground states of the dispersive hamiltonian and the full rabi model .
unfortunately , the degree of squeezing @xmath80 is relatively low even for the largest values of the light - matter coupling strength @xmath7 for which the dispersive approximation holds . in this section
we outline a strategy to significantly improve the amount of squeezing of the oscillator mode .
the basis of our protocol is a scheme detailed in ref .
@xcite , which uses sudden changes in the frequency of a harmonic oscillator to generate arbitrarily strong squeezing of the oscillator state .
we show that this scheme can be directly realised within the dispersive hamiltonian , using the interaction of the qubit with the field mode to create the required frequency shifts . the use of a dispersively coupled qubit distinguishes our proposal from that of ref .
@xcite , in which the repeated frequency shift protocol of ref .
@xcite was shown to produce squeezed states in a nonlinear superconducting oscillator . to begin with we re - express the hamiltonian in terms of position and momentum coordinates for mode @xmath1 :
@xmath110 so that the dispersive hamiltonian of eq . can be written as @xmath111 within each qubit subspace the hamiltonian of the field takes the form of a harmonic oscillator of unit mass and shifted oscillation frequency @xmath112 .
the lowering and raising operators associated with the frequency - shifted potentials are defined as @xmath113 note that the modes defined by @xmath114 are not independent : both sets of operators depend on the same underlying position and momentum coordinates , so that @xmath115 \neq 0 $ ] .
rather , the frequency - shifted mode operators are related to the original oscillator mode by @xmath116 .
now we imagine a scenario in which the system is initialized in the ground state of the dispersive hamiltonian @xmath117 . at time @xmath118 the state of the qubit @xmath119 is flipped suddenly : @xmath120 the orthogonality of the qubit states @xmath46 and @xmath45 implies that prior to @xmath118 the effective hamiltonian of the field mode is @xmath121 , whereas for @xmath122 the mode evolves under the effective hamiltonian @xmath123 . in other words ,
since the qubit state determines the frequency shift of the mode , flipping the qubit results in a sudden change in the frequency of the harmonic oscillator .
it has been shown that a suitably timed sequence of sudden frequency changes is capable of generating arbitrarily large squeezing of a field mode @xcite . in our coupled light - matter system , the dispersive interaction together with qubit flips provides the mechanism for changing the frequency of the field mode .
for the sake of completeness we briefly summarize the main steps of the protocol developed in ref .
@xcite within the context of our system .
* we assume that for @xmath124 the system is prepared in the ground state of the dispersive hamiltonian @xmath125 , where @xmath126 is defined as the ground state of the frequency - shifted oscillator potential @xmath127 .
* at @xmath118 the qubit is suddenly flipped to its excited state @xmath45 .
immediately following the qubit flip , although the expectation values of @xmath128 and @xmath129 for the field remain unchanged , the field state is squeezed _ relative to the new potential _ @xmath130 .
* the joint state evolves under the hamiltonian @xmath131 for a time duration @xmath132 .
* after the time delay @xmath132 , the qubit is suddenly flipped back to its ground state @xmath46 .
this creates a second sudden frequency jump from @xmath133 , following which the state of the field is squeezed with respect to the potential @xmath127 . * the first cycle of the protocol finishes with allowing the joint state to evolve under the hamiltonian @xmath134 for a duration @xmath135 .
* by carefully choosing @xmath136 and repeating the above steps @xmath137 times very strong squeezing of the field mode @xmath1 can be generated .
these steps are illustrated in fig .
[ sqzng1cycle ] .
as long as we are working within the dispersive hamiltonian , the above protocol is directly analogous to the harmonic oscillator with time - dependent frequency envisioned by janszky and adam @xcite .
following their heisenberg - picture analysis , we arrive at the following expressions for the time - evolved operators @xmath138 after one cycle of the protocol : @xmath139 the operators for the @xmath140 potential immediately following the second qubit flip are given by @xmath141 \end{split } \\
\begin{split } \hat{a}_{-}^{\dagger}(\delta t_{+ } ) & = \cos ( \omega_{+}\delta t_{+})\hat{a}_{-}^{\dagger}(-0 ) \\ & \quad + i \sin ( \omega_{+}\delta t_{+})[u_{+}^{2}\hat{a}_{-}^{\dagger}(-0)+u_{-}^{2 } \hat{a}_{-}(-0 ) ] , \end{split}\end{aligned}\ ] ] where @xmath142 are the initial creation and destruction operators before the start of the protocol and @xmath143 .
the time evolved operators @xmath144 depend crucially on the choice of @xmath132 .
if @xmath145 no squeezing is generated by the protocol .
maximal squeezing is obtained when @xmath146 . likewise , if the protocol is to be further repeated it is critical to choose @xmath147 . from now on we therefore fix @xmath148 .
the time evolved position and momentum quadratures defined with respect to the mode with frequency @xmath140 then become @xmath149 the corresponding variances of the quadrature operators are given by @xmath150 which clearly illustrates that the protocol decreases quantum noise beyond the standard quantum limit in the momentum quadrature at the expense of increased fluctuations in the position quadrature of the field mode .
this protocol can be repeated to produce even greater squeezing : after @xmath137 cycles the variances are given by @xmath151 and @xmath152 . unlike the proposal in ref .
@xcite , the time - dependent frequency changes in our system are achieved via an effective interaction with a qubit .
it makes sense , therefore , to consider the squeezing relative to the bare cavity mode with frequency @xmath3 and lowering and raising operators @xmath153 .
following one cycle of the protocol , the time - evolved quadrature operators for the bare cavity mode @xmath154 are related to the operators for the @xmath155 mode by @xmath156 after @xmath137 cycles of the protocol the variances in the position and momentum quadratures of the bare cavity mode are given by @xmath157 and @xmath158 , respectively . in the calculations that follow the degree of squeezing @xmath80
is always computed with respect to the bare cavity mode .
figure [ compdispjumpprot ] compares the degree of squeezing present in the ground state @xmath159 with that obtained after one cycle of the protocol , as a function of the coupling strength @xmath160 .
it can be clearly seen that the protocol using sudden frequency flips can significantly increase the degree of squeezing of the field mode over that naturally present in the ground state . .
for comparison the squeezing present in the ground state @xmath159 is also shown ( solid ) .
other physical parameters include : @xmath161.,scaledwidth=49.0% ] the foregoing discussion has been based on the dispersive hamiltonian and its eigenstates
. however , in our system the dispersive hamiltonian arises as an approximation to the full rabi hamiltonian . as the squeezing protocol
involves manipulating the state of the system , working with the rabi hamiltonian itself will give different results than working with the dispersive hamiltonian . in order to analyze the outcome of the protocol using the rabi hamiltonian , it is useful to work in the schrdinger picture rather than the heisenberg picture used previously . in the schrdinger picture ,
one cycle of the protocol based on the dispersive hamiltonian as described above results in the state @xmath162 the protocol may be carried out similarly using the rabi hamiltonian ; however , this does result in some deviation from the ideal case using the dispersive hamiltonian .
for simplicity , we assume that the initial state for the rabi protocol is the ground state of the rabi hamiltonian . as is evident from the approximate solution in eq . , this state is not an eigenstate of the qubit operator @xmath163 ; rather than being a separable state of qubit and field with the qubit in @xmath46 , the rabi ground state has a component along @xmath45 .
however , provided that the parameter regime is chosen such that the dispersive approximation holds , the @xmath45 component is small and will not have a significant negative impact on the degree of squeezing produced by the protocol .
as in the dispersive case , the first step of the protocol is to flip the state of the qubit by applying the @xmath164 operator . of course , this does not result in a state that is purely along @xmath45 , but again the error induced is small .
the flipped state is now allowed to evolve under the full rabi hamiltonian for a time @xmath165 . following this evolution ,
the qubit is flipped again and the state evolves , still under the full rabi hamiltonian , for a time @xmath166 .
the evolution time intervals @xmath136 remain the same as in the dispersive case , to a good approximation .
this is because the evolution time is related to the differences in energy eigenvalues , which are effectively the same in the dispersive and rabi cases provided that the parameters are chosen suitably .
the resultant state after one cycle of the protocol is then given by @xmath167 of course , the protocol may then be repeated multiple times in a similar fashion .
figure [ fiddisprabi ] shows the results of numerical calculations of the degree of squeezing @xmath80 that is produced as a function of the number of cycles @xmath137 for both the dispersive and rabi cases .
it is clear that in both cases the degree of squeezing increases linearly with the number of cycles .
remarkably , fig .
[ fiddisprabi ] shows that the protocol using the rabi hamiltonian produces considerably more squeezing per cycle ( approximately twice as much , for these parameters ) than the `` ideal '' dispersive case .
this is particularly surprising since the ground state of the rabi model shows less squeezing than the dispersive ground state , which can be attributed to the higher variance of the mode state associated with the @xmath45 state of the qubit as discussed at the end of sec .
[ sec : disp ] .
calculations show that the @xmath46 component of the state is squeezed much more strongly after one cycle of the rabi protocol than predictions based on the dispersive analysis would suggest .
the variance in @xmath168 of the @xmath45 state is also reduced by the squeezing protocol , albeit not below the standard quantum limit of @xmath169 .
however , the overall increase in squeezing in the rabi case can be attributed to the very large degree of squeezing of the @xmath46 component . clearly ,
although the dispersive theory gives good predictions for the energies and the degree of squeezing in the ground state in the rabi model , it is not particularly good at predicting the results from the rabi squeezing protocol .
this counterintuitive result highlights the need to be cautious when applying the dispersive approximation in situations where the state of the system is being manipulated @xcite . for the dispersive ( red , triangles ) and
rabi ( black , circles ) versions of the squeezing protocol , as a function of the number of cycles @xmath137 .
other physical parameters include : @xmath170.,scaledwidth=49.0% ]
in a realistic setting the protocol presented in the previous section will suffer from losses and imperfections .
qubit dephasing and relaxation , cavity losses , timing jitter , and the inability to instantaneously flip the state of the qubit will all affect the outcome of the protocol . we briefly discuss each of these sources of error with an eye toward circuit qed experiments , but their relative contributions will depend on the particular experimental implementation . in the dispersive version of the protocol ,
the qubit is always in one of its eigenstates , meaning that the protocol is unaffected by dephasing and only the energy relaxation time @xmath171 needs to be considered .
the rabi case is a little more complicated and qubit dephasing may contribute to noise .
however , it is important to note that the qubit need only remain coherent over the evolution time interval @xmath136 rather than throughout the full @xmath137 cycles of the protocol . to take some numbers relevant to superconducting circuit qed experiments , choosing a cavity frequency @xmath172 ghz and a coupling strength @xmath173 gives @xmath136 on the order of nanoseconds . given that superconducting qubits are now routinely achieving relaxation and dephasing times of several tens of microseconds @xcite , it is clear that qubit losses are not expected to be a limiting factor for the protocol or the degree of squeezing that can be achieved .
the loss of photons from the cavity has a more severe effect . to incorporate photon losses in the squeezing protocol
we use a standard master equation approach @xcite , assuming that during the time intervals @xmath136 the field mode couples to a zero temperature reservoir with damping rate @xmath174 .
figure 6 shows the results of a master equation simulation with @xmath175 . in the presence of photon loss the degree of squeezing
is no longer linear in the number of cycles @xmath137 ; the additional squeezing generated by another round of the protocol begins to saturate as @xmath137 increases .
however , a substantial increase in squeezing over that present in the initial state can still be obtained .
it is worth noting that this value of @xmath174 corresponds to a cavity q of about 150 , which is two orders of magnitude less than q - factors routinely achieved for microwave resonators in circuit qed @xcite and similar to the value recently measured for a qubit - coupled nanomechanical resonator @xcite .
the protocol also requires carefully designed time delays to ensure maximum squeezing of the field mode . in order to model the effect of inaccuracies in timing , we add a random offset @xmath176 , chosen from a normal distribution with zero mean and standard deviation @xmath177 , to each time delay @xmath136
. the average degree of squeezing of the field mode @xmath178 is evaluated by taking an ensemble average over multiple runs of the @xmath137-cycle protocol .
figure [ errorwithtime ] shows the average degree of squeezing @xmath178 when photon losses and random time delay are incorporated in our protocol . when the error in timing is on the order of 1 % , the timing jitter has little effect on the degree of squeezing .
however , larger timing errors ( on the order of 10 % ) dramatically reduce the degree of squeezing produced .
hence the ability to control the timing of qubit flips to a reasonably precise degree will be important for experimental implementations of our protocol . produced by @xmath137 cycles of the rabi protocol .
the three lower curves show the effect of photon loss together with three different degrees of timing imperfections : photon loss alone with no timing error ( brown , triangles ) , photon loss plus random timing errors chosen from a gaussian distribution of width @xmath179 ( yellow , crosses ) , and photon loss plus random timing errors with distribution width @xmath180 ( cyan , open circles ) . for comparison , the ideal case is also shown ( black , filled dots ) .
the photon loss rate is @xmath175 ; other physical parameters are given by @xmath181.,scaledwidth=50.0% ] the foregoing calculations have relied on the assumption of instantaneous flips of the qubit state , which produce sudden changes in the frequency of the oscillator .
a sudden frequency change is of course an idealization for a finite but small switching time @xmath182 .
as long as @xmath183 , a sudden frequency change from @xmath184 is a good approximation @xcite .
it should be pointed out that @xmath185 ; therefore the requirement for very small values of @xmath182 can be relaxed by decreasing the light - matter coupling @xmath7 and increasing the number of cycles @xmath137 to achieve a similar degree of squeezing of the cavity field .
moreover , modifications to the sudden frequency shift scheme of ref .
@xcite have been discussed in the literature .
the case of a sinusoidal frequency modulation was studied in ref .
@xcite , and a fourier - modified janszky - adam scheme for improved nonadiabatic generation of squeezed photons was proposed in ref .
while these schemes are not as efficient as the original sudden jump protocol , they are less demanding from an experimental standpoint and still produce significant squeezing of the cavity field .
in this work we have theoretically explored the potential for generating strong squeezing of a boson field mode interacting with a two - level system in the dispersive regime , without making the rwa .
although the ground state of this hybrid quantum system exhibits squeezing of the field , the degree of squeezing is not large .
however , the dispersive frequency shift allows the frequency of the cavity mode to be changed by flipping the state of the qubit . a protocol based on a series of suitably timed sudden frequency jumps
can be used to produce an arbitrarily large degree of squeezing in the absence of noise . even in the presence of a realistic level of noise and experimental imperfections ,
the degree of squeezing produced by this protocol can be significant .
one possible advantage of this method of squeezing generation compared to the use of a parametric amplifier and/or other nonlinearity based methods is that under our protocol the degree of squeezing and the time of generation are directly controlled by the number of frequency shifts applied . c. rigetti
_ et al . _ , _ phys . rev .
b _ * 86 * , 100506(r ) ( 2012 ) . r. barends _ et al .
_ , _ phys .
lett . _ * 111 * , 080502 ( 2013 ) .
j. braumller , _ appl .
lett . _ * 108 * , 032601 ( 2016 ) .
_ , arxiv : 1508.06299v2
. s. l. braunstein , and p. van loock , _ rev .
phys . _ * 77 * , 513 ( 2005 ) .
h. p. yuen and j. h. shapiro , _ ieee trans .
inf . theory _ * 24 * , 657 ( 1978 ) .
r. e. slusher _ et al .
_ , _ phys . rev
. lett . _ * 55 * , 2409 ( 1985 ) .
h. chen , y - m .
zhang , and x. wu , _ phys .
b _ * 40 * , 11326 ( 1989 ) .
t. sandu , v. chihaia , and w.p .
kirk , _ j. lumin .
_ * 101 * , 101 ( 2003 ) .
s. ashhab , and f. nori , _ phys .
a _ * 81 * , 042311 ( 2010 ) . s. ashhab , _ phys .
a _ * 87 * , 013826 ( 2013 ) .
l. c. g. govia , and f. k. wilhelm , _ phys .
a _ * 93 * , 012316 ( 2016 ) .
_ , _ nature _ * 479 * 376 ( 2011 ) .
s. matsuo , t. fujii , n. hatakenaka , _ physica b _ * 468 * 57 ( 2015 ) . | we present a protocol to generate a large degree of squeezing of a boson ( light ) field mode strongly coupled to a two - level system in the dispersive regime .
our protocol exploits the strong dispersive coupling to introduce a time dependent frequency change of the boson field . with an appropriately timed sequence of sudden frequency changes ,
the quantum noise fluctuations in one quadrature of the field can be reduced well below the standard quantum limit , with a correspondingly increased uncertainty in the orthogonal quadrature . even in the presence of realistic noise and imperfections
, the protocol should be capable of generating substantial squeezing with present experimental capabilities . |
the nature of the dark matter ( dm ) is a fundamental missing piece of the dark puzzle of the universe , and represents one of the hottest challenges facing particle physics and cosmology today .
the latest observational data @xcite prefer a flat universe with a @xmath1 mass fraction of dm , seeded with gaussian , scale dependent adiabatic perturbations .
the bulk of dm is believed to be `` cold '' , i.e. composed by weakly interacting particles that were non - relativistic at the epoch of decoupling .
the most popular candidate for cold dark matter ( cdm ) is the lightest supersymmetric particle ( lsp ) which , in most supersymmetry breaking scenarios , is the neutralino @xmath2 .
neutralinos are spin-@xmath3 majorana fermions , linear combination of the neutral gauge bosons and neutral higgs doublet spartners , @xmath4 . in the popular sugra or sugra - like models , where gaugino - universality is required ,
its mass is constrained by accelerator searches and by theoretical considerations of thermal freeze - out to lie in the range @xmath5 @xcite .
neutralinos decouple at a temperature that is roughly @xmath6 , hence they behave like cdm . if @xmath7-parity is conserved , neutralinos are stable and can change their cosmological abundance only through annihilation , which implies that the resulting production of detectable matter , antimatter , neutrinos and photons is enhanced in high density regions like the center of virialized dm halos . at low energies ,
direct detection terrestrial experiments are being performed to measure the energy deposited by elastic neutralino - nucleon scattering @xcite . besides direct searches , indirect detection of dm through its annihilation products can be implemented @xcite , which are necessary for energies greater than @xmath8 .
most indirect searches rely on detection of antimatter @xcite and @xmath0-rays produced in the milky way halo @xcite or in galactic sub - structures @xcite ; the possibility of looking at extragalactic sources has also been suggested @xcite .
unfortunately , our partial knowledge of those astrophysical processes that govern the assembly of dm concentrations did not allow to make firm predictions on dm detectability .
in particular , the central structure of the dm halos and the prominence of small scale structures within the smooth dm halo are far from being well determined .
the purpose of this work is to explore the possibility of extending dm searches to extragalactic sources in our local group of galaxies ( lg ) and to make robust predictions about their actual detectability with ground - based experiments once model uncertainties are taken into account .
supersymmetric lsp are not the only particles which can produce annihilation signals ( e.g. @xcite ) , yet in this work we focus on the contribution of neutralinos only .
the outline of the paper is as follows : in section ii we discuss the theoretical framework and present the predictions from the smooth halo models . in section iii and iv
we show how predictions change when the presence of supermassive black holes ( smbh ) and that of a population sub - galactic halos are taken into account .
section v deals with the actual detectability of lg sources and galactic sub - halos by ground - based detectors .
our conclusions are presented in section vi .
in a phenomenological approach which considers the lsp as a dm candidate , the expected photon flux from neutralino annihilation is given by : @xmath9 \frac{\vev{\sigma_{a } v}}{2 m^2_\chi } \int_{0}^{r_{max } } \frac{\rho_{\chi}^2 ( r ) } { { d^2 } } r^2 dr \ \ .
& & \label{flusso}\end{aligned}\ ] ] in this formula , the dependence of the flux on particle physics inputs and on cosmological inputs are factorized .
+ _ particle physics . _
the first sum in brackets represents a @xmath0-line , i.e. the 1-loop mediated process @xmath10 or @xmath11 . since neutralinos move at galactic speed @xcite , their annihilation occurs at rest and the outgoing photons carry an energy equal ( @xmath12 final state ) or very close ( @xmath13 final state ) to the neutralino mass . for the corresponding @xmath0-lines we have @xmath14 and @xmath15 monochromatic final state photons . branching ratios
@xmath16 imply that the photon flux is dominated by the continuum emission rather than by the @xmath0-lines .
the continuum emission is given by the second sum of eq .
( [ flusso ] ) running over all the tree - level final states . at the tree - level neutralinos annihilate into fermions , gauge bosons , higgs particles and gluons .
decay and/or hadronization in @xmath17 give a continuum spectrum of @xmath0-photons emerging from the @xmath17 decay .
depending on the annihilation channel , the continuum differential energy spectrum can be parametrized as @xmath18 , where @xmath19 , @xmath20 and @xmath21 .
these spectra have been calculated in @xcite using a pythia monte carlo simulation . branching ratio @xmath22 calculations for each process involved in the sum and the evaluation of the total annihilation cross section @xmath23 depend on the assumed supersymmetric model .
here we consider as a unique annihilation channel the one into w bosons , @xmath24 , with parameters @xmath25 and @xmath26 in the above mentioned @xmath27 formula .
we also use the constraint to the value of the thermally - averaged annihilation cross - section @xmath23 given by the neutralino number density at the freeze - out epoch @xcite : @xmath28 where @xmath29 is the hubble constant , @xmath30 , in units of @xmath31 and @xmath32 is the average density of neutralinos in units of critical density , @xmath33 . unless otherwise specified , throughout the paper we consider neutralinos with a mass of @xmath34 and cross section @xmath35 .
+ _ cosmology .
_ the last term in eq .
( [ flusso ] ) describes the geometry of the problem and assumes that the dm is concentrated in a single , spherical dm halo of radius @xmath36 and density profile @xmath37 located at distance @xmath38 from the observer .
the integral can be easily generalized to any dm distribution once that the dm density along the line of sight , @xmath37 , is specified .
an accurate evaluation of this integral should account for the fact that dm halos are triaxial rather than spherical objects @xcite .
deviations from sphericity do not change appreciably the @xmath0-ray flux from lg objects within the typical acceptance angle of the detectors ( @xmath39 ) and would only enhance the galactic flux by a modest @xmath40 @xcite and thus will not be considered here .
the shapes of the halo density profiles are poorly constrained by observations , especially near the center .
indeed , recent measurements of the dwarf galaxy rotation curves , as well as of the intracluster medium and of the gravitational lensing , could not discriminate between dark halos with constant density core and @xmath41 cusps @xcite .
similarly , analytical arguments did not yet lead to a definitive , unique prediction @xcite .
the most stringent constraints on the shape of dm halos are currently obtained from n - body simulations .
although the resolution in numerical experiments is still an issue @xcite , advances in numerical techniques have allowed to discover that dm halos have universal density shapes @xmath42 .
while singular isothermal halos characterized by a steep @xmath43 cusp are ruled out , current numerical experiments do not allow to discriminate among the moore profile @xcite characterized by @xmath44 and the shallower nfw profile @xcite with @xmath45 ( or slightly shallower ones @xcite ) .
in this paper we assume that the actual shape of the dm density profiles is bracketed by the moore and the nfw models and thus we can obtain a fair estimate of current model uncertainties by considering both of them in our analysis .
both models have central cusps where the neutralino annihilation rate is greatly enhanced .
the scale radius , @xmath46 , and the scale density , @xmath47 , can be fixed by observations ( the virial mass of the halo or its rotation velocity ) and theoretical considerations that allow to determine the concentration parameter @xmath48 ( where the virial radius , @xmath49 , is defined as the radius within which the halo average density is @xmath50 ) .
since the largest @xmath0-ray fluxes are expected from nearby objects , we have considered only the 44 nearest lg galaxies and their parent halos .
table [ tab1 ] lists the masses , positions ( from @xcite ) and virial radii of the galaxies which are more relevant for our analysis , along with their scale radii and scale densities computed assuming a moore profile .
their concentration parameters , @xmath51 and @xmath52 , have been computed according to @xcite assuming cdm power spectrum with a shape parameter @xmath53 , normalized to @xmath54 . in our analysis
we have also included the giant elliptical galaxy m87 at the center of the virgo cluster whose predicted annihilation flux was already computed by @xcite .
the mass , distance and the virial radius of m87 , listed in table [ tab1 ] , are taken from @xcite .
+ since the intensity of the @xmath0-ray flux is proportional to @xmath55 , great attention must be paid in modeling the central region of the halo .
the fact that the moore model predicts a divergent flux from the halo center implies that there must exist a minimum radius , @xmath56 , within which the self - annihilation rate @xmath57 equals the dynamical time @xmath58 , where @xmath59 is the mean halo density within @xmath56 @xcite .
we have used this prescription to determine @xmath56 for all our 44 halos and model their dm distribution with a nfw or a moore profile with a small constant core @xmath60 .
the resulting profiles are @xmath61 for the nfw case , and @xmath62 } , \ \ \ r > r_{min } \ \ \ \nonumber \\
\nonumber \rho_\chi(r ) & = & \rho_\chi^{moore}(r_{min } ) , \ \ \ r \leq r_{min } \\\end{aligned}\ ] ] for the moore case .
[ fig1 ] shows an aitoff projection of the angular position of these lg galaxies in galactic coordinates .
the size of each point is proportional to the @xmath0-ray flux from moore models , measured within a viewing angle of @xmath63 .
-ray flux emitted by a host dm halo with a moore profile within a viewing angle of @xmath63 from the halo center .
, width=302,height=226 ] as shown by @xcite , high energy @xmath0-ray flux from nearby extragalactic halos with a moore density profile can be larger than the @xmath0-ray foreground produced by neutralino annihilation in our galaxy .
this is not an entirely trivial point as distant extragalactic objects are not resolved within the typical @xmath63 angular resolution of the @xmath0-rays detectors . to check whether our conclusions still hold for shallower density profiles ,
we have computed the @xmath0-ray flux using both moore and nfw models for our lg halos and compared it to the galactic foreground . in this work
all fluxes are computed above @xmath64 within an angle @xmath65 .
note that none of the particle physics assumptions affect the ratio between galactic and extragalactic fluxes that we will focus on .
[ fig2 ] shows the smooth @xmath0-ray galactic foreground for a moore ( continuous line ) and a nfw ( dotted line ) profile as a function of @xmath66 , the angle from the galactic center ( gc ) .
filled triangles represent the @xmath0-fluxes from those lg galaxies which are brighter than the galactic foreground and from m87 , assuming a moore profile .
the fluxes from the small and large magellanic clouds ( smc , lmc ) , m31 and , to a lesser extent , m87 are above the galactic level . in some cases , like m33 and sagittarius , extragalactic and galactic contributions are comparable .
similar considerations are valid for the case of a nfw profile ( filled dots ) , despite of the significant reduction of the extragalactic fluxes that cause m33 and sagittarius to be dimmer than the galactic foreground .
.[tab1 ] input parameters for the @xmath0-ray flux prediction are shown for our galaxy , m87 and for the lg galaxies brighter than the galactic foreground .
masses and distances are taken from @xcite .
virial radii were calculated following the prescription given in the text .
scale radii and scale densities were computed assuming a moore profile . [
cols="^,^,^,^,^,^ " , ] in all _ dd _ models considered , the flux from m31 is increased by the presence of sub - halos while the flux from lmc decreases .
this discrepancy reflects the different effect of the tidal field that , as we have seen , disrupts all the sub - halos near the halo center and thus decreases the flux from the nearby lmc , while the flux from m31 , which is produced in a broader region , receives a significant contribution from the nearby , foreground sub - halos in m31 .
varying @xmath67 in the allowed range does not change the fluxes appreciably , while increasing the core radius of the sub - halo distribution decreases the m31 flux significantly but does not change significantly the flux received from the nearer lmc .
finally , neglecting the tidal field ( _ d0 _ model ) has the rather obvious effect of increasing the flux from both extragalactic objects . enhancing the tidal disruption effect ( _ d5 _ model )
generally decrease the annihilation flux , apart from the case of m31-nfw , in which the flux is enhanced by the combined effect of the halo distance and the large dm concentration .
in this section we explore the actual detectability of the @xmath0-ray fluxes from both the extragalactic sources and the galactic foreground .
satellite borne experiments , such as glast @xcite , have very small effective areas at energies @xmath68 and thus are inefficient in revealing high energy photons from neutralino annihilation .
the need for larger collecting areas can only be met by those ground - based experiments designed to detect the products of generic atmospheric showers .
these detectors were not originally meant to explore exotic physics , yet recent advances in technology have allowed to reach unprecedented high sensitivities and they can now be used to detect high energy photons from neutralino annihilation . + _ @xmath69erenkov detectors . _
electrons generated in showers emit @xmath70erenkov light during their passage through the atmosphere .
it reaches the ground in the form of a @xmath71 meters wide light pool .
@xmath70erenkov detectors have good angular and energetic resolution , operate with a duty cycle of about 10% and have a very good hadronic background rejection .
their small angle of view ( @xmath72 ) makes them suitable for the observations of point sources like lg galaxies . currently operating @xmath69erenkov telescopes
such whipple @xcite and hegra @xcite do not have enough sensitivity to detect faint extragalactic sources .
planned active @xmath70erenkov telescopes ( act ) such as hess @xcite , magic @xcite and veritas @xcite with their lower energy threshold ( @xmath73 ) and greater sensitivities make them suitable for our purposes .
_ large field of view arrays .
_ unconventional air shower arrays at energies @xmath74 tev constitute viable alternative to @xmath69erenkov telescopes .
unlike standard air shower arrays , these detectors can reconstruct lower energy showers by dense sampling of the shower particles .
this can be achieved by operating the detector at very high altitude in order to approach the maximum size development of low energy showers and by using a full coverage layer of counters .
operating large field of view arrays like argo @xcite and milagro @xcite have worse angular resolution and background rejection than act detectors , but are able to explore the whole sky with a duty cycle of about 100 % .
+ due to the complementarity of acts and large field of view arrays we explore and compare the capabilities of these detectors of revealing the @xmath0-ray flux from neutralino annihilation in extragalactic objects @xcite .
all analyses , including our own , have shown that the best place to look for dm annihilation signals is the gc , and indeed presently available and future detectors should be able to detect such @xmath0-ray emission
. however , most of them are ( or will be ) located in the northern hemisphere where the gc can only be seen for short periods and at uncomfortably large zenith angles . in fig .
[ fig6 ] the angular positions of our extragalactic objects ( filled dots ) are shown in galactic coordinates , superimposed to the areas that can be seen with a zenith angle smaller than @xmath75 from various experimental sites ( shaded areas ) . in the following we estimate the detectability of @xmath0-ray fluxes from either external lg galaxies or sub - halos within our galaxy . at the argo site ( upper panel ) ,
hess site ( lower left ) and an hypothetical antarctic site ( lower right ) .
terrestrial longitude and latitude of each sites are indicated above each panel .
, width=302,height=264 ] in the following we will focus on the detectability of m31 since it is the extragalactic object with the largest @xmath0-ray flux visible from the northern hemisphere . the detectability of the continuum flux from m31 is computed by comparing the number of @xmath0 events expected from this source to the fluctuations of background events . this ratio , which we call @xmath76 in the following , is given by : @xmath77 where @xmath78 is the time during which the source is seen with zenith angle @xmath79 and @xmath80 is the fraction of signal events within the optimal solid angle @xmath81 corresponding to the angular resolution of the instrument .
the effective detection areas @xmath82 for electromagnetic and hadronic induced showers are defined as the detection efficiency times the geometrical detection area . for the case of a large field of view array we have considered @xmath83 , depending on the @xmath0-photon energy , while for a @xmath70erenkov apparatus we have assumed a conservative effective area of @xmath84 .
note that while the latter can be increased by adding together more @xmath70erenkov telescopes , the former is intrinsically limited by the number of hits reaching the ground and can not be much greater than our fiducial value .
finally we have assumed an angular resolution @xmath85 and hadron - photon identification efficiencies of @xmath86 for the case of a large field of view array and @xmath87 and @xmath88 for the @xmath70erenkov apparatus .
these values are appropriate for the energy range of interest
. + _ backgrounds .
_ in our analysis the m31 flux is much smaller than electron , hadron and diffuse @xmath0-ray backgrounds that must then be taken into account .
we have considered the following values for the background levels : @xmath89 for the hadronic background @xcite , @xmath90 for the electron background @xcite , and @xmath91 for the diffuse extragalactic gamma emission , as extrapolated from egret data at lower energies @xcite .
+ the galactic diffuse emission along m31 line of sight @xcite turned out to be negligible with respect to the annihilation signal and has been neglected . + in fig .
[ fig7 ] the various backgrounds ( dashed lines ) are compared to the integrated energy spectra of @xmath0-photons from dm annihilation in m31 , computed assuming a moore dm profile and a population of sub - halos unaffected by tidal effects .
the various curves refer to different neutralino masses , as specified by the labels .
filled dots represent the expected @xmath0-line fluxes .
+ -line contribution ( filled dots ) are shown .
the relevant backgrounds ( dashed lines ) are shown for comparison.,width=302,height=264 ] fig .
[ fig8 ] compares the expected integrated @xmath0 flux from m31 ( dot - dashed curve ) to the @xmath92 sensitivity curves for the large field of view array ( upper line ) and the @xmath70erenkov telescope ( lower line ) .
the @xmath92 detection curves assume 1 year of data taking for the large field of view array and 20 days pointing for the @xmath70erenkov telescope .
+ the results show that annihilation signals from m31 , which are too faint for large field of view arrays , are well within reach of next generation @xmath70erenkov telescopes . sensitivity curves for a high altitude full coverage air shower array ( upper line ) and for a @xmath70erenkov cell ( lower line ) ,
computed using the parameters specified in the text .
the curve shows the same flux from m31 displayed in fig .
6 for the case of a neutralino mass of 1 tev.,width=302,height=264 ] the only way of detecting signals from annihilation in galactic sub - halos is through blind searches since model predictions specify the sub - halo distribution but not their precise spatial positions . with this respect ,
the use of a large field of view array has the advantage of collecting signals from a wide area of the sky , while acts would only allow to survey a much smaller region . to estimate the chances of detecting a @xmath0-ray signal with either detectors we performed ideal observations with both instruments and computed the significance of an annihilation signal using eq .
( [ sensitivity ] ) .
we restrict our observations to all directions with @xmath93 if @xmath94 around the gc and @xmath95 elsewhere provided that they are within a zenith angle of @xmath96 .
these constraints allow us to ignore the galactic diffuse photon contribution to the background emission and the effects of the atmospheric depth . for the case of a large field of view array we have chosen the argo experimental site ( @xmath97 e , @xmath98 n ) in tibet and considered the case of a 365 days observation with a detector having an angular resolution of @xmath63 .
each point in fig .
[ fig9 ] represents the significance of a @xmath0-ray signal detection in such experiment , plotted as a function of @xmath66 , the angle from the gc .
each filled dot accounts for the annihilation flux received from all galactic sub - halos with the same @xmath66 that have entered the detector field of view during the observation .
the galactic signal in absence of sub - halos is also plotted for reference ( open dots ) .
the significance detection level is very small at all @xmath66 but in a couple of cases where a very nearby sub - halo crosses the line of sight .
the results show that it would be impossible , using a large field of view array , to identify the galactic sub - halo origin of any detected @xmath0-ray flux .
+ for the case of an act we have chosen the veritas site ( @xmath99 w , @xmath100 n ) in arizona . in this case we have considered the results of a 20 days `` stacking '' observation in which an instrument with an angular resolution of 0.1 degrees points toward different directions characterized by the same value of @xmath66 . the results of this experiment are represented by the triangles in fig .
since each point represents a 20 days observations we have sampled only a few angles @xmath66 . although the significance of the act detections is higher than in the case of a large field of view array , the serendipitous nature of the @xmath0-ray sources combined with the small extension of the surveyed area make it very hard to detect the annihilation signal from galactic sub - halos using @xmath70erenkov telescopes .
+ we then conclude that only satellite borne experiments like glast will have the chance to detect @xmath0-ray photons from galactic sub - halos and to study their spatial distribution @xcite .
-ray signal from galactic sub - halos as a function of the angle of view from the gc , for a large field of view array ( dots ) and for a @xmath70erenkov telescope ( triangles ) .
the considered sub - halo distribution is not affected by any tidal effect.,width=302,height=264 ]
in this work we have computed the expected @xmath0-ray flux from dm annihilation in the nearest lg galaxies , under the assumption that the bulk of dm is constituted by neutralinos . predicted fluxes depend on a number of assumptions like the shape of the dm halo density profiles , the possible existence and the distribution of a population of sub - galactic halos and the dynamical influence of smbhs located at the halo centers .
uncertainties in these theoretical models that propagates into flux predictions have been evaluated by systematically exploring the space of the currently accepted model parameters .
it turns out that , in all cases considered , the fluxes from the small and the large magellanic clouds , m31 and m87 are well above the level of the @xmath0-ray annihilation galactic foreground
. uncertainties in the predicted galactic emission are quite large ( they vary by a factor @xmath71 ) and mainly result from the poorly known modeling of the dm sub - structures within our galaxy .
this scatter actually increases when considering more exotic possibilities that would further boost up the galactic foreground like the existence of caustics in the dm distribution @xcite or the presence of a population of mini sub - halos allowed by the very small cutoff mass in the cdm spectra @xcite .
neutralino annihilation in sub - halos contributes less to @xmath0-ray flux from extragalactic sources , resulting in a much smaller ( by a factor of @xmath101 ) model uncertainties and more robust flux predictions . in section
ii we have assumed that the inner slope of the dm density profile @xmath102 is in the range @xmath103 following the indications of numerical experiments rather than using observational constraints .
the reason was that current observations do not provide a self - consistent scenario .
indeed , recent determinations based on the x - ray properties of the intercluster medium seem to indicate a steep dm density profile @xmath104 , provided that clusters with disturbed x - ray surface brightness are removed from the sample @xcite . on the other hand ,
the rotation curves measured in a sample of lsbs by @xcite are consistent , on average , with a shallower density profile @xmath105 .
however , when the full extent of the rotation curves is taken into account rather than the inner region only , then the recovered density profiles lie between the nfw and the moore ones @xcite .
another study of high resolution @xmath106 rotation curves for dwarf and lsb @xcite has ruled out neither a profiles as steep as @xmath45 nor a constant density core .
finally , indications for shallow profiles ( @xmath107 ) were found by combining stellar dynamics and strong lensing data in a sample of brightest cluster galaxies @xcite , and from microlensing events toward the gc ( @xmath108 , in the hypothesis that the mw halo is spherically symmetric @xcite ) . however , estimates based on weak gravitational lensing in x - ray luminous clusters indicate a much steeper profile with @xmath109 @xcite .
+ the previous examples show that the current observations neither support nor reject profiles as shallow as @xmath110 or 0 . therefore , while we regard the nfw and the moore profiles as the most likely cases based on n - body results , we also want to keep an eye on the shallow profile case . for this purpose
we have estimated the expected @xmath0-ray flux from both m31 and lmc in the case of shallow profiles . if @xmath110 then the m31 flux is @xmath111 time smaller than in the nfw case and decreases by a further factor of 10 with a flat core @xmath112 .
results for the lmc are very similar .
therefore , if the slope of the density profile would turn out to be as shallow as @xmath110 or more , then no ground - based experiments would have the chance of detecting extragalactic @xmath0-ray annihilation fluxes .
+ in the hypothesis of both a nfw and a moore dm density profile we have estimated the possibility of detecting the annihilation signals from galactic sub - halos and extragalactic sources using those ground based instruments sensitive enough to detect @xmath0-photons in the @xmath73 - few @xmath113 energy band , such as acts or large field of view arrays .
we have found that the expected fluxes from galactic sub - halos are too faint to be detected by ground - based observatories .
their existence and distribution could instead be probed by next generation satellite borne experiments such as glast . on the other hand
, ground - based experiments should be able to detect a few extragalactic objects such as lmc and m31 .
in particular , we have been focusing on m31 which turned out to be the brightest extragalactic object visible in most experimental sites .
we have shown that next generation @xmath70erenkov telescopes should be capable of revealing the @xmath0-ray flux from m31 at a significance level @xmath114 in a 20 days pointing observation .
these results have been obtained using a quite conservative observational set up and theoretical modeling .
increasing the number of @xmath70erenkov cells ( as already planned for some future experiments ) or the serendipitous superposition of a nearby galactic sub - halo would certainly increase the chance of detecting an annihilation signal along the m31 direction .
the capability of detecting extragalactic signals is not shared by large shower arrays , which due to their lower sensitivity will not be able to detect such a faint source .
it is worth stressing that the gc does remain the best place to consider for detecting dm annihilation signatures .
unfortunately , and unlike m31 , it is not visible from the northern hemisphere where most of the ground - based detectors are ( or will be ) located . finally , it is worth stressing that our predictions assume that dm is made of neutralinos .
however , the natural factorization of particle physics and cosmology in the problem addressed in this paper makes straightforward to extend our approach to other kinds of weakly interacting dm candidates .
we thank m. de vincenzi , n. fornengo , p. gondolo , s. matarrese and b. moore for useful discussions and suggestions , and j. navarro and e. romano - diaz for having provided numerical codes and graphic tools used in this work .
99 d. n. spergel et al . , astrophys . j. suppl .
* 148 * , 175 ( 2003 ) .
j. ellis , et al .
d. * 62 * , 075010 ( 2000 ) .
g. jungman , et al .
, phys . rep . *
267 * , 195 ( 1996 ) .
n. fornengo , nucl.phys.proc.suppl .
* 95 * , 221 ( 2001 ) .
j. l. feng , et .
al , phys .
b * 482 * , 388 ( 2000 ) .
l. bergstrm , nucl.phys.proc.suppl . * 81 * , 22 , ( 2000 ) .
j. l. feng , et al .
, phys . rev . d. * 63 * , 045024 ( 2001 ) .
a. bottino , et al .
d. * 58 * , 123503 ( 1998 ) .
f. donato , et al .
d. * 62 * , 043003 ( 2000 ) .
m. urban , et al .
b * 293 * , 149 ( 1992 ) . v. berezinsky , et al .
b * 325 * , 136 ( 1994 ) .
l. bergstrm , et al .
, astroparticle phys . * 9 * , 137 ( 1998 ) . c. calcaneo - roldan and b. moore , phys .
d , * 62 * , 1230 ( 2000 ) .
r. aloisio , p. blasi , a. olinto , astro - ph/0206036 .
p. ullio , et al .
d * 66 * , 123502 ( 2002 ) .
j. taylor and j. silk , mnras . , * 339 * , 505 ( 2003 )
. l. bergstrm , j. edsj , p. ullio , phys .
lett . * 87 * , 1301 ( 2001 ) . c. tyler , phys .
d * 66 * , 023509 ( 2002 ) .
e. a. baltz , et al .
d * 61 * , 023514 ( 2000 ) .
d. fargion , et al .
, astropart.phys .
* 12 * , 307 , ( 2000 ) .
k. griest and m. kamionkowski , phys . rep . *
333 - 334 * , 167 ( 2000 ) .
b. moore , et al .
d * 64 * , 063508 ( 2001 ) .
f. stoher , et al .
, astro - ph/0307026 .
f. van den bosch and r. swaters , mnras .
, * 325 * , 1017 ( 2001 ) .
a. nusser and r. sheth , mnras . ,
* 303 * , 685 ( 1999 ) .
k. subramanian , astrophys .
j. , * 538 * , 528 ( 2000 ) .
a. dekel , et al . , astro - ph/0205448 .
c. power et al . , mnras . ,
* 338 * , 14 ( 2003 ) .
b. moore , et al .
, astrophys .
* 524 * , l19 ( 1999 ) .
j. f. navarro , et al .
, astrophys .
j. , * 490 * , 493 ( 1997 ) .
m. mateo , annual review of astronomy and astrophysics , * 36 * , 435 ( 1998 )
. + http://www.ast.cam.ac.uk/mike/local_members.html .
v.r . eke , et al .
, astrophys.j .
* 554 * , 114 ( 2001 ) .
d. mclaughlin , astrophys .
j. lett . ,
* 512 * , l9 ( 1999 ) .
v. berezinsky , et al .
b * 294 * , 221 ( 1992 ) .
p. gondolo and j. silk , phys .
. lett . * 83 * , 1719 ( 1999 ) .
p. ullio , h zhao , m. kamionkowski , phys .
d , * 64 * , 043504 ( 2001 ) .
m. begelman , et al .
, nature * 287 * , 307 ( 1980 ) .
p. c. peters , phys .
b * 136 * , 1224 ( 1964 ) .
d. merritt and f. cruz , astrophys .
j. * 551 * , l41 ( 2001 ) .
m. milosavljevi and d. merritt , astrophys .
j. , * 563 * , 34 ( 2001 ) .
m. volonteri , p. madau , f. haardt , astrophys .
j. * 593 * , 661 ( 2003 ) .
d. merritt , et al .
* 88 * , 191301 ( 2002 ) .
d. merritt , l. ferrarese , c. joseph . ,
, * 293 * , 1116 ( 2001 ) .
r. genzel et al .
mnras , * 317 * , 348 ( 2000 ) .
a. morselli , et al .
, in proc . of the 32nd rencontres
de moriond , ( 1997 ) .
j. h. buckley , et al .
, in proc . of `` towards a major atmospheric @xmath70erenkov detector iii '' , ( 1994 )
. v. fonseca , et al .
, nucl.phys.proc.suppl . , * 28a * , 409 , ( 1992 ) .
f. a. aharonian , et al . , astroparticle physics * 6 * , 343 ( 1997 ) . c. baixeras , et al .
, nucl.phys.proc.suppl . ,
* 114 * , 247 , ( 2003 ) .
t. c. weekes et al .
, in proc .
of the 25th icrc , * 5 * , 173 ( 1997 ) .
m. de vincenzi , et al .
, in proc .
of the 25th icrc , * 5 * , 265 ( 1997 ) .
r. atkins , et al .
, in proc .
of the 27th icrc , ( 2001 ) .
l. pieri and e. branchini , in proc . of the 28th icrc , ( 2003 ) , astro - ph/0307042 .
t. gaisse , et al .
, in proc . of the 27th icrc , ( 2001 )
. m. s. longair , high energy astrophysics , cambridge university press , ( 1992 ) .
p. sreekumar , et al .
, apj * 494 * , 523 , ( 1998 ) . s. f. shandarin and ya .
b. zeldovich , phys .
* 52 * , 1488 ( 1984 ) .
v. berezinsky , et al . , astro - ph/0301551 | under the hypothesis of a dark matter composed by supersymmetric particles like neutralinos , we investigate the possibility that their annihilation in the halos of nearby galaxies could produce detectable fluxes of @xmath0-photons .
expected fluxes depend on several , poorly known quantities such as the density profiles of dark matter halos , the existence and prominence of central density cusps and the presence of a population of sub - halos .
we find that , for all reasonable choices of dark matter halo models , the intensity of the @xmath0-ray flux from some of the nearest extragalactic objects , like m31 , is comparable or higher than the diffuse galactic foreground . we show that next generation ground - based experiments could have the sensitivity to reveal such fluxes which could help us unveiling the nature of dark matter particles . |
soft matter systems driven far from equilibrium by a shear flow manifest striking properties as a result of their sensitivity to external fields @xcite . for instance , hydrodynamics may either enhance or suppress coarsening processes @xcite .
coupling with the flow can also lead to the emergence of new , shear - induced phases that are exclusively out - of - equilibrium structures @xcite , whereas spatiotemporal oscillations and rheochaos are observed in shear - banding systems @xcite . despite substantial progress ,
a fundamental understanding of complex fluids under shear remains a challenging question .
the first reason lies in the nature of the coupling between structure and flow . since the local structure of soft materials
is readily reorganized by an external flow , it has in turn a significant impact on the flow itself .
the complexity of this feedback mechanism makes that most theoretical studies are based to a variable extent on phenomenological models @xcite .
the second point is that our understanding of nonequilibrium steady states ( ness ) is yet far from complete .
recently , there have been several attempts to construct a unified framework to describe equilibrium and nonequilibrium phenomena @xcite . in particular , fluctuations
theorems @xcite or extended fluctuation - dissipation relations @xcite have been suggested for complex fluids under shear .
but these relations still remain at the conceptual level and the link with experimental observables
power spectra , correlation functions has not been clarified yet . in this paper , we investigate the statistical properties of a liquid - liquid interface in a stationary flow .
we argue that this system is sophisticated enough to capture the relevant features of ness but remains simple enough to be addressed analytically .
applications of this issue might be expected for instance in microfluidics , since interfacial phenomena are inevitably enhanced as the size of the system is reduced .
still , the question of interface fluctuations under shear has received only little attention so far . from a fundamental viewpoint ,
equilibrium properties of liquid - liquid interfaces are now well established @xcite , and experimental as well as numerical studies have validated the capillary wave model down to almost molecular scales @xcite .
this topic has experienced a renewed interest in recent years with the discovery that phase - separated colloid - polymer mixtures may have an extremely low interfacial tension @xcite .
interface fluctuations can then be analyzed in real time and space using elementary video - microscopy techniques .
the versatility of this method has allowed derks and collaborators to study the statistical properties of an interface exposed to a shear flow @xcite .
they found that the coupling with the flow leads to a strong reduction of thermal fluctuations , while the correlation length increases .
but this second point is in disagreement with recent monte carlo simulations of a driven ising model @xcite , therefore raising the fundamental question of what features of interfaces under shear are actually universal .
statistical fluctuations of an interface are driven by the random forces that spontaneously occur in the bulk @xcite . to describe nonequilibrium properties ,
the main issue is thus to properly account for the coupling between the bulk and the interface .
this can be achieved on the basis of fluctuating hydrodynamics ( fh ) @xcite .
indeed , fh has been successfully applied to various ness situations , and experimental validations have been obtained , e.g. , for fluids in a temperature gradient @xcite .
the purpose of this paper is to apply fh to an interface under shear .
we shall proceed as follow . in sec .
[ fluct ] , we explain how an equation of motion for the interface can be obtained from fh .
we show that the distortion of capillary waves by the flow leads to a mode - coupling equation that is discussed in sec .
this allows us to extract ness properties under a stationary flow in sec .
in particular , we find that the fluctuations are smoothed out by the flow .
the results are then discussed in sec.[disc ] in the light of recent experimental and numerical data available in the literature .
finally , we conclude the paper with a short summary of our results . for the sake of clarity , details of the algebra
are presented in apps .
[ app1][app4 ] .
we first set up a hydrodynamic theory to account for the coupling between the surface and the bulk . following the usual hypothesis of capillary wave theory
, we assume the existence of an intrinsic interface separating two immiscible fluids . for moderate deformations around the @xmath0 horizontal plane ,
the position of the interface can be described by a single valued function @xmath1 . along this article , properties of the upper ( resp .
lower ) fluid are labeled with the subscript @xmath2 ( resp .
@xmath3 ) .
each phase is characterized by its mass density @xmath4 and its viscosity @xmath5 .
we also define @xmath6 the mean viscosity and @xmath7 the mass density difference .
the surface is further characterized by the interfacial tension @xmath8 and the capillary length @xmath9 , with @xmath10 the gravitational acceleration .
the system is schematically drawn in fig .
[ schema ] .
the average position of the interface is @xmath11 .
it is confined between to walls , the thickness of each fluid layer being @xmath12 and @xmath13 with @xmath14 .
a planar couette flow is induced by moving the walls of the shear cell at constant velocity along the @xmath15 direction .
we define @xmath15 , @xmath16 and @xmath17 respectively as the velocity , vorticity and velocity gradient directions . assuming that the no - slip condition applies on the walls of the cell , the fluid velocity @xmath18 satisfies @xmath19 and @xmath20 .
the total shear rate is then @xmath21 where we define @xmath22 the shear rate in each phase . without loss of generality , we assume in the following that @xmath23 , @xmath12 , @xmath24 and @xmath13 are chosen so that the plane of zero shear coincides with the average position of the interface .
other situations can be deduced thanks to a galilean transformation . for usual fluids at room temperature ,
thermal fluctuations occur in the overdamped regime of capillary waves . in order to describe nonequilibrium effects
, the analysis is performed within the framework of fluctuating hydrodynamics @xcite .
the starting point is the stochastic version of the stokes equation @xmath25 with @xmath18 the velocity field , @xmath26 the pressure , and @xmath27 .
( [ stokes ] ) is solved together with the incompressibility condition @xmath28 thermal fluctuations are accounted for through the random part of the stress tensor @xmath29 .
its components @xmath30 ( with @xmath31 , @xmath16 , or @xmath17 ) are stochastic forces that stem from the microscopic degrees of freedom of the fluids .
close to equilibrium their correlations are given by the fluctuation - dissipation theorem , but such a relation is not expected to hold beyond the regime of linear response . here however , we shall take advantage of the separation of time scales between the collective modes under study the fluctuations of the interface and the molecular scales of the heat bath the fluid constituants .
the relaxation of an interface is characterized by the capillary time @xmath32 ; it ranges from milliseconds for usual interfaces ( @xmath33 mn / m ) to a few seconds for ultra - soft interfaces ( @xmath34 mn / m ) . on the other hand , a particle of fluid on either side of the interface diffuses over its own diameter @xmath35 on a time - scale @xmath36 .
this time scale is at most of the order of @xmath37 s for particulate fluids with @xmath38 nm @xcite , but it is several orders of magnitude smaller for molecular fluids .
since the relevant regime considered in this work corresponds to @xmath39 , the applied shear rate is too small to significantly affect the thermal motion of individual particles . in this small peclet number limit ( @xmath40 ) ,
the statistics of the heat bath is not affected by the flow and we can assume that the stochastic variables have zero mean value and correlations given by @xcite @xmath41 with @xmath42 the boltzmann s constant and @xmath43 the temperature .
is normal to the interface and is pointing towards the upper fluid.,width=302 ] eqs .
( [ stokes ] ) and ( [ incomp ] ) are then solved above and below the interface , and the solutions matched with the appropriate boundary conditions .
the latter have to be enforced at the instantaneous location of the interface @xmath1 .
explicitly , we need to express the continuity of the velocity @xmath44_h = 0 \ , \label{velocity}\ ] ] as well as the continuity of the stress @xmath45_h \cdot \mathbf{n } = \sigma \ , \mathbf{n } \left ( \bm{\nabla } \cdot \mathbf{n } \right ) \ , \label{stress}\ ] ] with the notation @xmath46_{z_0 } = f(z_0^+)-f(z_0 ^ -)$ ] , the limit being taken respectively from above and from below . in eq .
( [ stress ] ) , @xmath47 is the _ total _ stress tensor @xcite .
the components of @xmath48 read @xmath49 , with @xmath50 ; the random part @xmath29 is defined in eq .
( [ stokes ] ) .
the unit vector @xmath51 is normal to the surface , pointing towards the upper fluid .
it depends on the local conformation of the interface @xcite @xmath52 finally , once the velocity field is fully characterized , an equation of motion is obtained thanks to the kinematic relation @xmath53 the velocity being evaluated at @xmath54 . in this equation , the parallel components of a vector field @xmath55 are defined as @xmath56 . for the sake of completeness
, the derivation of the boundary conditions ( [ stress ] ) and ( [ kinematic ] ) is reminded in appendix [ app1 ] .
it has been appreciated for a long time that interfacial fluctuations can be distorted by an external flow @xcite .
the description of this effect requires to go beyond the linear analysis generally used to derive the dispersion relation for capillary waves . to proceed
, we follow a recursive scheme that has proved its worth , e.g. , in the context of polymer - membrane interactions @xcite .
we assume that the deformation can be written @xmath57 , with @xmath58 .
the dimensionless parameter that governs this small - gradient expansion is @xmath59 .
the velocity and pressure fields are then expressed as @xmath60 the lowest - order term is simply the couette flow solution for a planar interface : @xmath61 if @xmath62 , and @xmath63 if @xmath64 . for @xmath65
, the linearity of the stokes equation implies that each term of the series obeys eq .
( [ stokes ] ) .
the coupling between successive orders originates from the boundary conditions .
indeed , the latter have to be enforced at the position of the interface .
it is therefore expected that the taylor expansion of the @xmath66-order field , when evaluated at @xmath67 , involves contributions of all orders @xmath68 . to be more explicit ,
consider for instance the velocity field .
up to second order , it is given by @xmath69 \nonumber \\ & + & \varepsilon^2 \left [ \mathbf{v}^{(2 ) } ( 0 ) + u \partial_z \mathbf{v}^{(1 ) } ( 0 ) + \frac{u^2}{2}\partial_z^2 \mathbf{v}^{(0 ) } ( 0 ) \right ] + \ldots \nonumber \end{aligned}\ ] ] [ note that we have only written the @xmath17-dependence of @xmath70 .
we proceed likewise for all fields in eq .
( [ velocity ] ) ( [ kinematic ] ) .
clearly , the taylor expansion of the boundary conditions involves several second - order contributions , making the algebra quite cumbersome .
moreover , the study of the fluctuations requires us to solve the full three - dimensional problem .
we thus defer the technical details to app .
[ app2 ] and [ app3 ] , and we now focus the discussion on the main outcomes . since the problem is invariant by translation parallel to the horizontal plane , it is natural to switch to the ( 2d ) fourier representation @xmath71h(\mathbf{r},t)$ ] , with @xmath72 and @xmath73 .
we also define the norm of the wave vector @xmath74 . after some algebra , we find that the relaxation of a fluctuation mode with wave vector @xmath75 follows a _ mode - coupling equation _
@xmath76 this equation constitutes the first main result of this paper .
it involves a number of contributions that we now discuss .
first , @xmath77 is the equilibrium relaxation time of the interface .
this result shows that there is no direct coupling at linear order ( at least in the viscous regime ) .
advection of the interface by the flow takes the form of a convolution between all modes with an effective shear rate @xmath78 note that the second - order term does not depend on the elastic properties of the interface . the special feature of eqs .
( [ mceq ] ) and ( [ gammaeff ] ) is that the effective shear rate @xmath79 felt by the interface differs from the applied shear rate @xmath80 defined in eq .
( [ gammatot ] ) .
although the latter is set by the geometry of the system , the former is a dynamical quantity in the sense that it depends on the viscosities of both fluids .
however , @xmath81 and @xmath80 can not be tuned independently since the continuity condition ( [ stress ] ) for tangential forces requires @xmath82 ( see app . [ app2 ] ) , and then @xmath83 this relation implies that the effective shear rate can in principle be tuned in the range @xmath84 by adjusting the experimental conditions .
it is only when @xmath85 that both shear rates coincide .
thermal fluctuations are accounted for through the white noise @xmath86 .
this contribution stems from the random part of the stress tensor ( see appendix [ app3 ] ) .
we find that @xmath86 simply follows the equilibrium distribution with zero mean value and @xmath87 even though the system is driven far from equilibrium , there is no coupling between the noise and the external flow ( at least up to @xmath88 ) @xcite . at this point
, it should be mentioned that corrections similar to eq .
( [ mceq ] ) have been suggested in the context of sheared smectic phases ( see @xcite and references therein ) , or in the description of coarsening under shear @xcite .
the argument commonly invoked is the following .
since the deformation @xmath89 is dragged along the @xmath15 direction by the external flow @xmath90 , the time derivative involved in the equation of motion can simply be replaced by the total derivative @xmath91 where @xmath92 is evaluated at @xmath54 .
of course if the slope of the velocity is continuous , i.e. if @xmath93 , there is no ambiguity : the total derivative then reads @xmath94 , yielding to the convolution integral in fourier space .
however , the situation is more subtle when @xmath95 . with the same argument ,
the velocity evaluated at @xmath54 would be @xmath96 if @xmath97 , whereas it would take the value @xmath98 if @xmath99 .
clearly , there is an inconsistency in the reasoning .
the only way to remove the indeterminacy is to follow the procedure detailed in appendix [ app2 ] .
we emphasize in particular that the various second - order contributions in eq .
( [ velocity ] ) ( [ kinematic ] ) are all equally important to give the perfectly symmetric relation eq .
( [ gammaeff ] ) .
ness properties of the interface are then extracted from the equation of motion ( [ mceq ] ) together with the noise correlations ( [ noisecorrel ] ) .
unfortunately , this class of nonlinear stochastic equation can not be solved exactly .
we are thus forced to restrict the discussion to moderate shear rates .
but this does _ not _ mean that @xmath100 should remain small , as might be expected from simple scaling arguments .
indeed , our study reveals that the parameter that actually governs the steady state properties of the system is the combination of the two dimensionless parameters @xmath101 depending on experimental conditions , @xmath102 can be small even if @xmath100 is not @xcite .
the analysis presented in the following is valid in the range @xmath103 .
( [ mceq ] ) is solved using a perturbation theory presented in appendix [ app4 ] .
we first discuss the steady - state correlation function @xmath104 defined as @xmath105 it is given at equilibrium by @xmath106 . under shear , the spectrum is modified according to @xmath107 \ , \label{sqshear}\ ] ] with @xmath108 the angle between the direction of shear @xmath109 and the wave vector @xmath75 .
the function @xmath110 depends only on the norm @xmath111 of the wave vector .
explicitly , we get @xmath112^{-1 } -\left[sf(s)\right]^{-1 } } { f(x)+f(s)+f ( \vert \mathbf{x}-\mathbf{s } \vert ) } } \ , \label{integral } \end{aligned}\ ] ] with @xmath113 , @xmath114 , @xmath115 the angle between @xmath116 and @xmath117 , and @xmath118 .
this integral can not be evaluated analytically ; the result of the numerical integration is presented in the inset of fig .
[ sdeq ] @xcite .
( [ sqshear ] ) is the second main result of this paper .
it shows that the coupling is maximum in the flow direction ( @xmath119 ) , while the spectrum is not affected in the vorticity direction ( @xmath120 ) .
as expected , the result is invariant by inversion symmetry @xmath121 or @xmath122 .
the correction @xmath123 is plotted in fig .
[ sdeq ] in dimensionless units .
it should be noticed that even though all the wave lengths are affected by the flow , the spectrum is mostly affected when @xmath124 .
the correction then vanishes like @xmath125 for larger values of @xmath126 , whereas it scales as @xmath127 in the limit @xmath128 .
, in the direction of shear @xmath129 and for @xmath130 .
in inset we show @xmath110 .
, width=302 ] the mean square displacement of the interface is then obtained from the sum over all modes @xmath131 .
notice that the integral that defines the roughness at equilibrium is actually divergent .
regularization is achieved by means of a microscopic cut - off @xmath35 so that @xmath132 . in ness , we find that the fluctuations are strongly reduced by the external flow @xmath133 \ .\ ] ] the correction is quadratic in the control parameter @xmath102 . in this expression , @xmath134 is a _
universal _ constant in the sense that it depends neither on the properties of the fluids nor of the elastic constants of the interface @xcite .
moreover , it is independent of the microscopic cut - off as soon as @xmath135 .
numerically , we get @xmath136 . for different values of the control parameter :
@xmath137 ( solid line ) , @xmath138 ( dot ) , @xmath139 ( dash - dot ) , and @xmath140 ( dash - dot - dot ) .
a ) in the direction of shear ( @xmath141 ) , the correlation length decreases .
b ) in the vorticity direction ( @xmath142 ) , the correlation length increases . , width=302 ] finally , we focus our attention on the ness correlation function @xmath143 , which is the fourier transform of the structure factor . as can be seen in eq .
( [ sqshear ] ) , the spectrum @xmath104 is not modified for the modes that are perpendicular to the flow ( _ i.e. _ wave vectors with @xmath144 ) .
however , performing the sum over all wave vectors gives rise , in direct space , to modifications _ even in the vorticity direction_. denoting @xmath145 the angle between the direction of shear @xmath109 and the position vector @xmath146 , the correlation function reads @xmath147 \ , \label{correl } \end{aligned}\ ] ] with @xmath148 the modified bessel function of the second kind .
the corrections can be written @xmath149 and @xmath150 , where for @xmath151 @xmath152 here , @xmath153 is the bessel function of the first kind . at equilibrium , @xmath154 is proportional to @xmath155 and behaves like @xmath156 $ ] at large separations @xmath157 .
in other words the capillary length is also the equilibrium correlation length .
in nonequilibrium situation , the correlation function presents a number of interesting features .
we first consider the direction of shear @xmath141 .
it clearly appears from fig .
[ correlfig].a ) that @xmath158 decays faster that its equilibrium counterpart .
this reflects a _ decrease _ of the correlation length as the control parameter @xmath102 is increased .
the situation is different in the vorticity direction @xmath159 , as shown on fig .
[ correlfig].b ) . in this case
the decay of @xmath158 is slower than the equilibrium correlation function , indicating an _ increase _ of the correlation length with increasing shear rate .
since we do not have an explicit functional form of the correlation function , it is difficult to be more quantitative regarding the correlation length .
the only conclusion that can be drawn from fig .
[ correlfig ] is that the decay of @xmath158 at large distances is not exponential anymore .
finally note that at intermediate angles ( @xmath160 ) , the correlation length can be either larger or smaller than @xmath161 depending on the shear rate .
complex fluids under shear is an challenging class of nonequilibrium systems .
the simple system considered in this work is a fluid interface exposed to a couette flow .
first , it was shown that the time evolution of the deformation satisfies a mode - coupling equation .
note that equations similar to eq .
( [ mceq ] ) have been suggested in the context of soft surfaces under shear ( see for instance @xcite ) . as a matter of fact
, it belongs to the general class of the kardar - parisi - zhang equation @xcite whose derivation is usually based on phenomenological grounds .
here however , eq . ( [ mceq ] )
has been _ rigorously _ derived from hydrodynamics with no other assumption that inertial effects can be neglected .
we now discuss our results in view of recent studies .
we have shown that interfacial fluctuations are strongly reduced by the flow . according to eq .
( [ hcarre ] ) , the reduction is governed by a universal parameter : we predict @xmath136 .
suppression of thermal capillary waves was indeed measured by derks and collaborators in a recent experiment @xcite .
this was achieved by using a phase - separated colloid - polymer mixture whose interface is characterized by a very low surface tension .
the authors have used two compositions but the dynamical parameters are only available for the first sample ( sample a , very close to the critical point ) .
the thickness of the fluid layers are @xmath162 @xmath163 m and @xmath164 @xmath163 m .
equilibrium parameters obtained from the correlation functions are @xmath165 nn / m , @xmath166 @xmath163 m , and @xmath167 s @xcite .
the viscosities extracted from the velocity profiles are @xmath168 mpa.s and @xmath169 mpa.s @xcite .
unfortunately , there are only two experimental points ( out of 6 ) that correspond to the condition @xmath170 . to fit the data for higher values of the control parameter
, we assume that eq .
( [ hcarre ] ) can be extended according to @xmath171 we then perform a linear regression of the quantity @xmath172 as a function of @xmath173 .
we find @xmath174 , almost 3 times the expected value .
the comparison between our theoretical predictions and the experimental measurements is shown on fig .
as can be seen the agreement is only qualitative , even though it is difficult to be really conclusive since most data are outside the validity range of eq .
( [ hcarre ] ) @xcite . .
the solid circles refer to the experiment of ref .
@xcite ( sample a ) .
the dotted and solid lines are the theoretical predictions of eqs .
( [ hcarre ] ) and ( [ hcarre2 ] ) , respectively , with @xmath175 .
the dashed line corresponds to the fit of eq .
( [ hcarre2 ] ) with @xmath174 .
the dash - dotted line is the model of derks _ et al . _
@xcite.,width=302 ] in ref .
@xcite , the suppression of thermal capillary waves is supported by a remarkably simple model .
the authors state that only the slow modes with @xmath176 are affected by the flow . solving this inequality defines to two wave vectors @xmath177 and @xmath178 .
it is then assumed that wave vectors with @xmath179 no longer contribute to the fluctuation spectrum , whereas the contribution from wave vectors outside this range is left unchanged .
their prediction of the interfacial roughness as a function of the applied shear rate @xmath80 is in very good agreement with experimental data see fig .
[ msd ] . still , we argue that the hydrodynamic analysis presented in this paper provides new insights into the problem .
in particular , we have shown that fluctuations are driven out of equilibrium by the effective shear rate @xmath79 rather than the applied shear rate @xmath80 . for sample a presented above , the ratio of effective to applied shear rate is @xmath180 . now
, suppose that the experiment is done with the same sample composition excepted that the thicknesses of the fluid layers are inverted to @xmath181 @xmath163 m and @xmath182 @xmath163 m . in this case , the ratio would be @xmath183 : for the same value of the applied shear rate , @xmath102 is thus expected to increase by 50 @xmath184 .
the ensuing difference regarding the mean square displacement should be substantial enough for the distinction between applied and effective shear rate to be experimentally observed .
this would provide a direct validation of our analysis .
finally , let us discuss the correlation functions .
our study reveals that the static structure factor is strongly modified in the flow direction , whereas fluctuation modes in the vorticity direction are not affected .
this result confirms similar conclusions obtained from molecular dynamic simulations @xcite .
coming back in real space , we find that the correlation length decreases in the direction of shear .
this conclusion agrees with monte carlo simulations of a driven interface @xcite .
physically , it would mean that the external flow acts as an effective potential whose strength should increase with the shear rate . however , these findings are in disagreement with the results of derks _ et al . _ that observe an increase of the correlation length @xcite .
note that experimental data are fitted using the equilibrium correlation function .
this procedure is mainly based on the short distance behavior of @xmath185 .
in contrast , our conclusions are drawn from the _ asymptotic _ behaviour of the correlation function .
it is thus possible that the discrepancy comes from how the correlation length is defined in practice .
other possibilities are discussed below .
the equation of motion ( [ mceq ] ) is a direct generalization of the dispersion relation for capillary waves .
in particular , it is based on the same assumptions : continuum hydrodynamics and small - gradient expansion .
the pertinent question is whether the apparent mismatch between theoretical predictions and experimental data can be the consequence of another mechanism @xcite .
in fact , several explanations are plausible .
firstly , we remark that the diameter of the colloids is of the order of @xmath186 nm .
the capillary length is thus only ten or twenty times larger than the `` microscopic '' cut - off , and it is not clear whether the separation of length scales assumed to get the universal quantities ( @xmath134 , @xmath187 , ) is really achieved for this sample . secondly , monte carlo simulations of smith _ et al .
_ reveals that the average _ width _ of the interface is significantly reduced by the flow @xcite . in the capillary wave theory
, it is assumed that the density profile behaves like a step function with a strictly vanishing width .
but it is known that physical parameters are directly related to the shape of the _ true _ density profile @xcite . for instance
, the surface tension is given by @xmath188 with @xmath189 the equilibrium profile and @xmath190 a constant proportional to the second moment of the intermolecular potential @xcite .
in the small but finite interfacial region , the density is varying rapidly so that the fluids are highly compressible . the density should therefore be modified by the flow and become a function of the applied shear rate @xmath191 .
a plausible consequence is that the bare surface tension might depend on @xmath80 as well . according to eq .
( [ defsigma ] ) , the thinning of the average profile observed in @xcite is expected to give rise to an increase of the surface tension , and then to an additional reduction of the fluctuations .
the latter assertion is certainly hypothetical but we argue that the measurement of the correlation function in the _ vorticity _ direction could be conclusive . indeed , an increase of the bare surface tension would lead to an increase of @xmath161 and thus of the ness correlation length .
but a modification of @xmath158 arising from @xmath192 is expected to be _ isotropic _ , whereas the hydrodynamic theory predicts a modification that is _
anisotropic_. this question clearly deserves to be reconsidered both from the experimental and theoretical viewpoint .
let us briefly summarize the main results of this paper .
\(i ) we have generalized the hydrodynamic derivation of the dispersion relation to include the effect of an applied shear rate on the dynamics of capillary waves .
( [ mceq ] ) reveals that the system is driven out of equilibrium by an effective shear rate that differs from the applied one .
the mode - coupling structure of the equation also shows that nonequilibrium fluctuations can not be described in terms of a simple renormalization of the interfacial tension .
\(ii ) the analysis of the mode - coupling equation shows that the fluctuations are smoothed out by the flow , in qualitative agreement with experiments and simulations .
we find that ness fluctuations can be described in terms of a single control parameter defined in eq .
( [ alpha ] ) .
conclusions regarding the height correlation function are more questionable . if the theory and the simulations predict a decrease of the correlation length in the direction of shear , the reverse is actually observed on the experimental side .
the discrepancies might be due to the finite thickness of the interfacial region and its response to the external flow .
molecular dynamics simulations of interfaces under shear should help to clarify this point in a near future . in conclusion ,
we have derived a hydrodynamic theory to account for the coupling between a shear flow and the fluctuations of an interface .
previous investigations on driven interfaces have revealed partial agreements but also significant discrepancies between experiments and numerical simulations .
although our work brings new insights into the issue , it seems that further investigations will be necessary to definitely reconcile experiment , theory and simulations .
ultimately , achievement of this program should allow to reach a deeper understanding of soft matter systems under shear .
the authors wish to thank d. aarts , d. derks , e. raphal , m. schmidt and a. wrger for stimulating discussions .
we are also indebted to d. derks for providing us with experimental data .
the institute of fundamental physics ( ipf ) bordeaux is acknowledged for financial support in the early stages of this work .
since the analysis presented in this work involves nonlinear contributions , particular care has to be given to the conditions at the interface . in this appendix , we first remind the derivation of the kinematic relation eq .
( [ kinematic ] ) .
we then focus on the stress continuity condition eq .
( [ stress ] ) .
it is convenient to represent the interface with the functional @xmath193 .
the unit vector @xmath194 normal to the surface is given by @xmath195 let us focus on a point @xmath196 that belongs to the interface .
it satisfies @xmath197 and then @xmath198 if @xmath199 is the velocity of the point @xmath200 , the infinitesimal displacement reads @xmath201 and therefore @xmath202 . together with eq .
( [ normalapp ] ) , we then obtain @xmath203 identifying the normal velocity of the interface @xmath204 with the normal velocity of the fluid @xmath205 , one finally gets @xmath206 here , we have also used the fact that @xmath207 . this relation
together with the definition ( [ normal ] ) directly leads to eq .
( [ kinematic ] ) . to derive the stress condition ( [ stress ] )
, we consider an infinitesimal surface element @xmath208 of the interface , bound by a closed contour @xmath209 .
the contour is chosen circular with radius @xmath210 .
we also define a small cylinder @xmath211 , centered on the contour @xmath209 and with height @xmath212 .
the forces acting on the volume element @xmath211 are the body force density @xmath213 , the tension force exerted along the perimeter @xmath209 , and the surface force exerted by the fluids above and below the cylinder .
we discuss separately each contribution .
we first discuss the restoring force exerted on the contour line @xmath209 .
we assume that the contour is travelled counterclockwise .
the ( local ) orthonormal basis associated with the line is denoted @xmath214 , with @xmath215 the tangent vector , @xmath117 the vector normal to @xmath209 but tangent to @xmath208 , and @xmath194 the vector normal to @xmath208 pointing towards the upper fluid . in this representation ,
the external force exerted on the contour @xmath209 reads @xmath216 with @xmath217 .
then , according to the stokes relation @xmath218 , one gets @xmath219 with the surface element @xmath220 . finally ,
using the general identity @xmath221 and because @xmath222 , we find @xmath223 the component @xmath224 of the stress tensor @xmath225 corresponds the @xmath226 component of the force ( per unit area ) acting on a surface element normal to the direction @xmath227 .
in particular , the force density exerted by a fluid on a surface element with outward normal @xmath228 is @xmath229 .
let us apply this definition to the elementary cylinder of volume @xmath211 .
the cylinder is sitting across the surface , the extension being @xmath210 on each side . regarding the upper surface of the cylinder , the stress exerted by fluid 1 is characterized by the ( outside ) normal vector @xmath230 so that the force density reads @xmath231 . similarly on the lower surface one
has @xmath232 and the stress is @xmath233 .
the total surface force is then @xmath234 the force balance on the volume element @xmath211 enclosing the surface @xmath208 defined by the contour @xmath209 then reads @xmath235 now if @xmath210 is the typical lengthscale of the volume element @xmath211 , then the surface contributions scale as @xmath236 whereas the acceleration and the body forces scale as @xmath237 .
hence in the limit @xmath238 the latter can be neglected and surface forces must balance .
however , since the surface element is arbitrary , the integrand must vanish identically so that one finally gets the desired result @xmath239 finally notice that the divergence of the normal vector reads in the monge representation @xcite @xmath240
the aim of this section is to derive the equation of motion ( [ mceq ] ) .
the fluctuation modes of the interface are expected to be deformed by the shear flow .
as we shall see , the description of this effect requires to go beyond the usual linear analysis .
we focus here on the relaxation dynamics in the absence of noise .
thermal fluctuations are treated in appendix [ app3 ] .
to begin with , we recall the stokes equation @xmath241 with @xmath242 the gravitational acceleration . in the small - gradient approximation
, we can express the deformation profile as @xmath243 with @xmath244 .
the dimensionless parameter that governs the small - gradient expansion is @xmath245 we then assume that the fields that satisfy the stokes equation can be expanded in powers of @xmath246 @xmath247 where @xmath248 stands for @xmath18 , @xmath48 or @xmath26 .
( for clarity reason , only the @xmath17-dependence of the fields is kept explicitly in this appendix ) .
the solution at order @xmath249 satisfies eq .
( [ stokesapp ] ) ; it corresponds to the couette flow solution for a flat interface .
the other terms of the expansion ( [ devapp ] ) are solution of the following equation @xmath250 at order @xmath251 , the solution in the absence of shear would lead to the usual dispersion relation for capillary waves . here , the calculations are carried out up to order @xmath252 . due to the linearity of eq .
( [ stokesnapp ] ) , the coupling between different orders only occurs through the boundary conditions at the interface .
this suggests to use a recursive method in order to solve the problem @xcite . for a flat interface
, the solution is the well - known solution for the couette flow : @xmath253 with @xmath2 if @xmath62 and @xmath3 if @xmath254 .
@xmath255 is a constant .
the stress tensor then reads @xmath256 in particular , the continuity relation ( [ stress ] ) for the tangential stress implies @xmath257 .
note that we have chosen @xmath12 , @xmath23 , @xmath13 and @xmath24 so that the plane of zero shear is the horizontal plane @xmath11 . at constant shear rate
, any other situation can be deduced thanks to a simple galilean transformation .
the first - order solution satisfies eq .
( [ stokesnapp ] ) with boundary conditions that involve the solution at order @xmath249
. actually , there are two kinds of conditions that needs to be enforced : continuity conditions at the interface and limit conditions far from the interface . strictly speaking
, the latter have to be expressed at the boundaries of the shear cell , @xmath258 or @xmath259 , where @xmath12 and @xmath13 are _ macroscopic _ lengths . in this work ,
we are interested on thermal fluctuations that develop on _ microscopic _ lengthscales ( i.e. , with wavelength much smaller than @xmath12 or @xmath13 ) .
we can therefore assume that for any field @xmath248 , the terms of the series ( [ devapp ] ) with @xmath65 vanish at infinity @xmath260 in the following , we first derive the boundary conditions at the interface and then solve the problem in fourier representation . to express the continuity conditions , let us define for a field @xmath248 the notation @xmath261_{z_0 } = f(z_0^+)-f(z_0 ^ -)\ ] ] with @xmath262 , the limit being taken for @xmath263 .
the first condition that needs to be enforced is the continuity of the velocity at the interface @xmath264_h= \mathbf{0 } \ .\ ] ] using the expansion ( [ devapp ] ) together with the taylor expansion of @xmath265 , we obtain at first order @xmath266_0 = \mathbf{0 } \ .\ ] ] from the result ( [ v0 ] ) we get @xmath267 , and finally @xmath268_0 & = & \left(\dot{\gamma}_2-\dot{\gamma}_1 \right ) u \ ,
\label{velxapp } \\ \left [ v_y^{(1 ) } \right]_0 & = & 0 \ , \label{velyapp } \\ \left [ v_z^{(1 ) } \right]_0 & = & 0 \ . \label{velzapp } \end{aligned}\ ] ] [ velocityapp ] next , we have to enforce the condition of stress continuity eq .
( [ stress ] ) . at first order
, we obtain @xmath269_0 = - \sigma \nabla^2 u \delta _ { \mu z } \ ] ] with @xmath270 or @xmath17 . following the result ( [ t0 ] ) for the stress tensor ,
we find @xmath271 the continuity conditions for the stress tensor then reads @xmath272_0 & = & 0 \ , \label{stressxapp } \\ \left[t_{yz}^{(1)}\right]_0 & = & 0 \ , \label{stressyapp } \\
\left[t_{z z}^{(1 ) } \right]_0 & = & \delta \rho g u - \sigma \nabla^2 u \ , \label{stresszapp } \end{aligned}\ ] ] [ stressapp ] with @xmath273 .
the solution of the problem is expressed using the two - dimensional fourier transform @xmath274f(\mathbf{r},z , t ) } \ , \\ f ( \mathbf{r},z , t)=\displaystyle{\int \frac{{\text{d}}^2 \mathbf{q}}{(2\pi)^2 } \exp [ -i\mathbf{q } \cdot \mathbf{r}]f(\mathbf{q},z , t ) } \ , \end{cases } \label{fourier}\ ] ] with @xmath72 and @xmath73 .
the difficulty of the algebra comes from the fact that we need to solve the full 3-dimensional problem .
it then appears judicious to define a new orthogonal coordinate system that would account for the symmetries of the problem . to this aim
, the vector fields are decomposed into their longitudinal , transverse and normal components @xcite .
this defines a new set of orthogonal unit vectors @xmath275 where @xmath276 is the unit vector parallel to @xmath75 , and @xmath277 the in - plane vector perpendicular to @xmath276 and @xmath278 .
these vectors are expressed in the cartesian basis @xmath279 as @xmath280 the velocity is then written as @xmath281 . inserting this expression in eq .
( [ stokesnapp ] ) leads to a system of differential equations for the fourier - transformed quantities @xmath282 [ stokesappfourier ] together with the divergenceless condition @xmath283 it is not difficult to show that the solution of eqs .
( [ stokesappfourier ] ) and ( [ incompressibility ] ) can be written [ solo1 ] @xmath284 @xmath285 , @xmath286 , and @xmath287 are the ( yet undetermined ) integration constants , the subscript @xmath288 taking the value @xmath2 if @xmath62 and @xmath3 if @xmath64 .
( there should be _ a priori _ no confusion between the subscript and the imaginary unit @xmath288 ) .
the solution for the longitudinal component follows directly from ( [ incompressibility ] ) .
note that the condition ( [ limit ] ) has already been taken into account in ( [ solo1 ] ) .
next , we have to express the boundary conditions eqs .
( [ velocityapp ] ) and ( [ stressapp ] ) in fourier space and in the new system of coordinates .
the condition ( [ velzapp ] ) for the normal component is straightforward @xmath289_0 = 0 \ , \label{vzq}\ ] ] where of course @xmath290 now stands for the fourier transformed quantity . then projecting conditions ( [ velxapp ] ) and
( [ velyapp ] ) for the parallel components onto the directions @xmath276 and @xmath277 respectively gives @xmath291_0 = \frac{q_x}{q } \left(\dot{\gamma}_2-\dot{\gamma}_1 \right ) u \ , \label{vlq}\ ] ] and @xmath292_0 = \frac{q_y}{q } \left(\dot{\gamma}_1-\dot{\gamma}_2 \right ) u \ .
\label{vtq}\ ] ] regarding the continuity of the stress , the condition ( [ stresszapp ] ) is now expressed as @xmath293_0 = \sigma \left(q^2 + l_c^{-2 } \right ) u \ , \label{szq}\ ] ] where we define the capillary length by @xmath294 .
finally , the projection of ( [ stressxapp ] ) and ( [ stressyapp ] ) leads to @xmath295_0 = 0 \ , \label{slq}\ ] ] and @xmath296_0 = 0 \ . \label{stq}\ ] ] with the boundary conditions ( [ vzq ] ) ( [ stq ] ) , we can now evaluate the 6 integration constants .
we find [ inconst1 ] @xmath297 with @xmath298 .
although the first - order solution already depends on the shear flow , it is claimed in the main body of the paper that the calculations have to be performed up to second order . to understand this point ,
consider the kinematic condition ( [ kinematic ] ) .
since @xmath299 , we can write at first order @xmath300 with @xmath301 . it appears from ( [ a11 ] ) that the shear rate is not involved in the equation of motion at linear order .
the calculations have thus to be extended to include the first nonlinear contribution .
the solution at second order satisfies the same set of linear differential equations ( [ stokesappfourier ] ) and ( [ incompressibility ] ) , so that it takes the form [ solo2 ] @xmath302 again , @xmath303 is deduced from the incompressibility condition ( [ incompressibility ] ) .
it appears that not all the integration constants are needed for our purpose .
indeed , let us express the kinematic condition ( [ kinematic ] ) up to second order .
in direct space , it reads @xmath304 where we have used the fact that @xmath305 .
note that the passage from ( [ kinematic ] ) to this latter equation leads to an indeterminacy .
indeed , although the velocity field @xmath18 is continuous at @xmath54 , this is not the case for each term of the expansion ( [ devapp ] ) at @xmath11 . in this appendix
, we thus assume that the limit @xmath306 is taken from above .
it can be checked that the same results are obtained if one would take the limit from below .
we now switch to fourier space , where the product of two functions @xmath307 is transformed into the convolution product @xmath308 we then have in fourier representation @xmath309 + \varepsilon^2v_z ^{(2)}(0 ) \ .\end{aligned}\ ] ] with the results of the preceding section , the equation of motion can now be written as @xmath310 + \varepsilon^2 a_1^{(2 ) } \ . \label{motionapp}\ ] ] therefore , only the integration constant @xmath311 is required to get the equation of motion . at second order ,
the continuity condition ( [ velocity ] ) for the @xmath17-component of the velocity is expressed in direct space as @xmath312_0 = 0 \ , \ ] ] leading to , in reciprocal space , @xmath313_0 = i \left ( \dot{\gamma}_2 - \dot{\gamma}_1 \right ) \big(u \ast q_x u \big ) \ . \label{vz2}\ ] ] this relation actually involves both @xmath311 and @xmath314 . to obtain one last equation
, we still need to express the normal stress condition @xmath315_h = - \sigma \nabla^2 h + \mathcal{o } \left ( h^3 \right ) \ .\ ] ] together with ( [ t0 ] ) and ( [ stressapp ] ) , it reads at order @xmath252 @xmath316_0 = 0 \ .\ ] ] this result is finally expressed in fourier representation .
the derivative of the stress tensor at first order is then obtained from eq .
( [ stokesn ] ) @xmath317 where we have used ( [ incompressibility ] ) to get the last equality . from the continuity relation ( [ slq ] ) , it is not difficult to obtain the condition @xmath318_0 = 0 \ , \ ] ] that leads to the simple relation @xmath319 inserting this result in eq .
( [ vz2 ] ) , we finally get @xmath320 if we factorize all second - order contributions in ( [ motionapp ] ) , we then arrive at the equation of motion @xmath321 coming back to the deformation field @xmath322 and using the fact that @xmath82 , we can write the equation of motion in its definitive form eq .
( [ mceq ] ) .
in this appendix , we consider the landau - lifshitz equation of linear fluctuating hydrodynamics @xcite . we thus need to adapt the perturbative scheme developed in appendix [ app2 ] to the stokes equation ( [ stokes ] ) including the random part of the stress tensor @xmath29 . to this aim
, it should be noticed that the noise correlations @xmath323 given in eq .
( [ randomten ] ) scale as @xmath324 and are therefore proportional to @xmath252 [ see eq .
( [ epsapp ] ) ] . from this viewpoint ,
the random part of the stress tensor can be identified as a _ first - order contribution_. we then follow the same recursive method as in appendix [ app2 ] .
there is no modification at order @xmath249 and we now discuss the first and second orders of the expansion ( [ devapp ] ) . at first order , the stochastic version of the stokes equation ( [ stokes ] )
reads , in the @xmath275 basis , @xmath325 [ fluctstokes ] together with the incompressibility condition @xmath326 as explained in the text , the fluctuations of the random forces in the bulk are not affected by the shear flow ( at least in the regime considered in this paper ) .
the correlations are then given by the fluctuation - dissipation theorem @xcite . in this representation , we have @xmath327 where the subscripts stand for @xmath328 , @xmath329 or @xmath17 . the solution of eqs .
( [ fluctstokes ] ) and ( [ fluctinc ] ) is then obtained using the following green s function identity @xcite @xmath330 \frac{1}{2q } e^{-q\vert z - z ' \vert } = - \delta ( z - z ' ) \ .\ ] ] regarding the boundary conditions , the stress tensor @xmath48 in ( [ velocityapp ] ) and ( [ stressapp ] ) has now to be replaced by the _ total _
stress tensor @xmath331 @xcite .
after some algebra , we obtain the evolution equation for the interface @xmath332 where the noise @xmath86 is given by @xmath333 together with the correlations ( [ fluctcor ] ) , we finally get the result announced in eq .
( [ noisecorrel ] ) .
note that eq .
( [ appcorphi ] ) generalizes the result of grant and desai for a liquid - gas interface ( in the absence of shear ) @xcite . at second order ,
the velocity field satisfies the following set of equations @xmath334 together with the continuity equation @xmath335 the noise enters the problem only through the boundary conditions and the first order contribution .
again , the boundary conditions are the same as in appendix [ app3 ] excepted that the stress tensor @xmath48 is replaced by the total stress tensor @xmath336 .
the calculations are pretty lengthy so that we only give the final results .
the equation of motion is obtained from the kinematic condition eq .
( [ kinematic ] ) . up to second order
, we find @xmath337 + \varphi - \mathbf{q } \cdot \big [ \left(\mathbf{q } \phi_1\right ) \ast h \big ] - \big ( \mathbf{q } \times \big [ \left(\mathbf{q } \phi_2\right ) \ast h \big ] \big ) \cdot \mathbf{e}_z \ , \ ] ] where @xmath338 and @xmath339 are defined as @xmath340 and @xmath341 it can be noticed that the additional terms involving @xmath338 and @xmath339 are not induced by the external flow . at equilibrium , these second - order contributions would actually renormalize the relaxation dynamics of the interface . in nonequilibrium situation , it is argued in the next appendix that those terms do not contribute to the fluctuation spectrum at the order considered in this work .
therefore , we do not to mention them in the main document .
this choice is made in order to improve the readability and to focus the discussion on the main outcomes .
eq . ( [ mceq ] ) is solved using a perturbation theory . to this aim ,
let us define the time fourier transform @xmath342 of any function @xmath343 as @xmath344f(\mathbf{q},t ) $ ] .
the mode - coupling equation can then be rewritten @xmath345 with @xmath346 .
the bare propagator @xmath347 is given by @xmath348 the solution of the problem is then obtained as the solution of a dyson equation .
the calculations are performed up to second order in the small parameter @xmath102 ( which is proportional to @xmath349 ) . using the shortland notation @xmath350 , we get from this expansion , the evaluation of the correlation function @xmath352 is pretty lengthy but presents no conceptual difficulty since the four - point correlation functions of the noise are given by the wick s theorem .
the result is then fourier - transformed back in time representation , leading to eqs .
( [ sqshear ] ) and ( [ integral ] ) .
the same scheme is used to treat the nonlinear noise contributions .
first , we need to specify the statistical properties of @xmath338 and @xmath339 . from ( [ fluctcor ] ) , it is not difficult to get @xmath353 , and for @xmath163 , @xmath354 or @xmath355 , @xmath356 and @xmath357 the solution is then expressed as a dyson equation similar to eq .
( [ appdyson ] ) .
although the expression of @xmath358 involves more terms , it can be shown that most of the additional contributions to @xmath359 actually vanish ( either because they involve an odd number of noise terms , or because of the vanishing cross - correlations between @xmath145 , @xmath338 and @xmath339 ) . as a matter of fact , additional shear - induced contributions are at least of order @xmath360 , and thus do not contribute to the modification of the spectrum at the order considered in this work .
still , the nonlinear noise _ does _ give rise to non - vanishing contributions at order @xmath361 .
but those terms are _ equilibrium _ contributions since they do not involve the shear rate .
consider for instance the mean square displacement : those terms would simply renormalize the equilibrium value @xmath362 .
this is why we have carefully plotted the _ ratio _
@xmath363 , that represents the modification of the fluctuations by the shear with respect to the _ equilibrium _ situation @xcite . as a matter of fact
, our analysis reveals that convolutions between deformation and noise of the form @xmath364 also occur in the equation of motion ( see appendix [ app3 ] ) .
these nonlinear terms are not coupled to the flow . at equilibrium
, they would actually renormalize the relaxation dynamics of the interface however , it is explained in appendix [ app4 ] that they do not contribute to the fluctuation spectrum at order @xmath361 , so that they will not be considered in the following . actually , the solution of eq .
( [ mceq ] ) can be extended to high values of the shear rate following for instance the self - consistent approach detailed in @xcite .
however , the results are very similar to those of the approxiation ( [ hcarre2 ] ) so that we do not present more details . | the steady state properties of an interface in a stationary couette flow are addressed within the framework of fluctuating hydrodynamics .
our study reveals that thermal fluctuations are driven out of equilibrium by an effective shear rate that differs from the applied one . in agreement with experiments ,
we find that the mean square displacement of the interface is strongly reduced by the flow .
we also show that nonequilibrium fluctuations present a certain degree of universality in the sense that all features of the fluids can be factorized into a single control parameter .
finally , the results are discussed in the light of recent experimental and numerical studies . |
understanding the dense packings of hard particles has yielded essential insights into the structure of materials @xcite , granular media @xcite , number theory @xcite , biology @xcite , and computer science @xcite .
this understanding has been hard won : hundreds of years can elapse between a conjecture and its proof .
this is best exemplified by sphere packing , for which a proof kepler s conjecture had not been found until 1998 @xcite .
recent experimental advances have allowed the development of ( nearly ) hard colloids that , for entropic reasons , manifest the densest sphere packing @xcite .
the attempt to obtain a deep understanding of liquid crystal mesophases has prompted study of the dense packing of anisotropic particles .
this has led to recent explorations of the packing of ellipsoids @xcite , polyhedra @xcite , and polygons @xcite .
one of the simplest regular polygons one can pack in two dimensions is the square . on the plane ,
the densest packing is , in this case , trivial a square lattice of squares .
monte carlo simulations of squares at finite pressure , however , have also found a tetratic phase @xcite , and experiments with hard colloidal squares have found , rather than the tetratic phase , a rhombic crystal having a different symmetry than the square @xcite .
even the dense packing of a finite number of squares can be more complicated than naive considerations would indicate in fact , the problem of packing squares in another square has been shown to be np - hard @xcite . indeed , the densest known packings can be quite complex @xcite when the number of squares is not a perfect square integer .
higher packing densities than that of a simple , square lattice with vacancies can be achieved through configurations in which some of the squares are rotated and shifted with respect to the square lattice @xcite . what is the effect of an external potential on packings of hard objects ?
one example of an external potential is fixed boundary conditions .
considerable effort has been devoted to understanding the densest packings of squares and circles in various domains .
another example with greater physical application is an external periodic potential .
if this potential is strong enough , the packing of the hard objects may be forced to be adopt the periodicity of the potential , and new packings are expected to arise . a limiting case of an imposed periodic potential is periodic boundary conditions . in this paper
we explore the densest configurations of hard squares in a torus that is , inside a larger square with periodic boundary conditions . even with the additional translation symmetry afforded by packing squares in a torus rather than in a square , the resulting dense packings in the torus can still be far from simple .
our results may have experimental relevance for hard square colloidal particles in a periodic potential imposed either by a substrate or an optical lattice .
as in many other mathematical packing problems , the strategy here is to search for the smallest area torus that can accommodate a fixed number of squares @xmath0 .
we use a combination of analytic and monte carlo simulated annealing techniques to accomplish this , and our results can be summarized as follows : we find that whenever @xmath0 can be expressed as the sum of two square integers
@xmath4 the densest possible configuration is a density - one packing with squares arranged in rows that are oriented at an angle of tan@xmath5 relative to the underlying torus . for other @xmath0
, we find a surprisingly rich collection of dense packing structures . for @xmath0
= 6,11,14 , and 27 , we believe that the densest possible packing is a commensurate bravais lattice packing with density @xmath1 and resembles a bricklayer pattern with periodic gaps . for @xmath6 12,21,22 and 23 , we find that the densest configurations are non - bravais lattice packings , including both regular lattices of holes and of rotated squares .
these results are summarized the table within section ii . in section
ii , we present a summary analysis of the various structures we found for @xmath0 up to 27 , including both commensurate bravais lattice solutions and non - bravais lattice solutions .
the packing motifs we found through analytic and numerical means are illustrated in this section via drawn figures as well as images generated by our numerical simulations . in section iii , we provide details of our numerical experiments for the hard square system .
section iv is a discussion of our results , including an analysis of the entropy of densest packings , and the rotational invariance of density - one packings as @xmath0 goes to infinity .
in this section we give an analytic treatment of square packings on a torus .
we first describe a set of solutions in which the squares lie on a bravais lattice , and then turn to more complicated cases .
the density - one solutions are optimal by construction , and all of the other solutions are conjectured to be optimal .
the numerical results of sec . [ sec : numerical ] guided us to the conjectured solutions , and the fact that long simulated annealing runs consistently produced these solutions gives us some confidence that they are optimal .
our conjectures for configurations containing @xmath7 squares are summarized in the table . here
we consider a class of bravais lattice configurations that includes all of the density - one packings and other solutions we have found for @xmath7 . in all of these packings , the squares are lined up in rows ; and for the purposes of this analysis , we assume that these rows are aligned along the x - axis .
thus , one of the primitive vectors of the lattice of squares is @xmath8 , and the second primitive vector is taken to have the form @xmath9 , with @xmath10 and @xmath11 .
note that the primitive vectors of the torus , @xmath12 and @xmath13 , need not be aligned with the primitive vectors of the squares .
the requirement that the squares pack periodically on the torus is equivalent to saying that the lattice of squares is commensurate with the larger square lattice of the torus .
that is , there exist integers @xmath14 , @xmath15 , @xmath16 and @xmath17 such that the torus primitive vectors @xmath12 and @xmath13 are given by @xmath18 in addition , we require that the torus primitive vectors are of equal length , @xmath19 these conditions are uniquely solved by @xmath20 the number of squares @xmath0 packed on the torus is the number of lattice points of the square lattice in a unit cell of the torus lattice @xmath21 and the areal density of the squares @xmath22 is given by @xmath23 there are two classes of density - one packings .
the first is the perfect square packing , for which @xmath24 , @xmath25 , @xmath26 and @xmath27 .
this simple packing is shown in fig .
[ fig : n9 ] for the case of @xmath28 .
note that , on the torus , each of the @xmath14 rows ( or columns , but not both ) may be arbitrarily displaced relative to the other rows ( columns ) without disturbing the density of the packing or its periodicity ; the perfect square packings thus have finite entropy .
[ fig : n9 ] there is a more general class of density - one packings , in which the lattice of squares may be tilted with respect to the primitive vectors of the torus . setting @xmath25 and @xmath24 in eq .
( [ eq : gap ] ) , we find @xmath29 , @xmath30 , and @xmath31 , the square lattice being oriented at an angle of @xmath32 relative to the torus lattice vectors
. clearly , these density - one , tilted square lattice solutions are optimal for all @xmath0 that are sums of two square integers .
note that the perfect square solution corresponds to the special case @xmath33 .
[ fig : bravais ] shows the case @xmath34 ( @xmath35 and @xmath36 ) .
we thus find that if n is the sum of two squares , there exists a density - one packing .
the converse of this is also true : @xmath0 is a sum of two squares for all density - one packings of squares in the torus . to prove this
, first note that every square in a density - one packing must have at least four other squares bordering it along a finite segment length , forcing all @xmath0 squares to share the same orientation .
now consider three squares in mutual contact with each other a configuration that must exist if the packing has no gaps .
two of those squares must be aligned in a row , as shown in fig .
[ fig : aligned ] . in order to eliminate gaps in the packing ,
these three squares define a set of rows that the entire packing must respect .
note that periodic boundary conditions allows us to draw the torus vectors so that they begin on the corner of a square and end on the corresponding corner of another square .
thus , @xmath15 and @xmath17 are both integers .
a right triangle can be constructed with @xmath15 as one side and @xmath37 as its hypotenuse ; another right triangle can be drawn with @xmath17 as its base , and @xmath38 as its hypotenuse .
( see fig .
[ fig : aligned ] ) .
these triangles are identical by inspection . @xmath39 and @xmath40 are therfore each equal to @xmath41 , via the pythogorean theorem . the simplest way to produce candidates for a densest packing for @xmath42 is to remove @xmath43 squares from a density - one packing ; indeed , our numerical results suggest that for several values of @xmath0 , the densest packing is a density - one packing with one missing square .
this is indicated in the comment column in the table using the notation @xmath44 or @xmath45 , depending on whether they are generated by removing @xmath46 square from @xmath0 a square integer , or a sum of two square integers respectively . as is demonstrated in fig .
[ fig : n7 ] for the case of @xmath47 , such vacancies allow for continuous displacement of other squares within a row , leading to a finite entropy for such configurations .
other examples include @xmath48 and @xmath49 .
next we consider bravais lattice solutions that have density less than one that is , packings with gap @xmath50 . because these solutions have rows that are shifted relative to one another ( @xmath51 ) , we call these `` gapped bricklayer configurations . ''
an example is shown in fig .
[ fig : gb](a ) . equations ( [ eq : gap ] ) allow us to ennumerate all gapped bricklayer configurations .
since the packing density of these configurations is @xmath52 , the highest packing density we can find within this class of configurations with density less than unity must have the form @xmath53 .
this requires that @xmath0 be one more than a sum of two squares .
the first several are gapped bricklayer configurations are seen for @xmath6 2,3,6,11,14,18,26 , and 27 .
based on the numerics we believe that for @xmath54 , 11 , 14 and 27 , the gapped bricklayer packing is the densest possible packing .
these are indicated in the table with the abbreviation `` gb '' in the comment column .
associated ( non - unique ) lattice vectors are shown in the rightmost columns of the table for the gb packings .
there are also gapped bricklayer solutions for density @xmath55 when @xmath0 is two more than a sum of two squares , though we have not found any candidate densest packing solutions of this form for @xmath7 .
unlike the bravais lattice packings with density one , different rows of the gapped bricklayer solutions have a fixed shift given by @xmath56 , where the denominator is the closest sum of two squares below @xmath0 . here
, we consider special cases suggested by our numerical simulations that do not correspond to bravais lattice packings .
note : if the comment column of the table simply repeats a value of @xmath0 , it indicates a special case for which the squares are not on a bravais lattice and for which there is not an obvious pattern that can be extrapolated easily to optimal packings for higher values of @xmath0 . the conjectured best packing for @xmath57 is shown in fig . [ fig : n22 ] .
this packing is in fact a gapped bricklayer configuration , except that the unit cell or brick is composed of two squares stacked in the @xmath58 direction ( the direction perpendicular to the rows ) .
the configuration is otherwise identical to the @xmath59 gapped bricklayer and has density @xmath60 .
the conjectured best configurations for @xmath62 and 23 are shown in figs .
[ fig : n12 ] and [ fig : n23 ] , respectively . in both cases
the motif can be described as a lattice of @xmath61 holes .
the torus lattice vectors , eq . [ eqn : ana ] with @xmath8 and @xmath64 , for @xmath62 are described by @xmath65 and the torus lattice vectors for @xmath2 are described by @xmath66 , and @xmath67 .
it is straightforward to verify that these motifs are in fact packings on the torus and have the density @xmath68 where @xmath43 is the number of holes in the unit cell . for @xmath62 , evidently @xmath69 and for @xmath2 , @xmath70 . the conjectured densest packing for @xmath71 , shown in fig . [ fig : n21 ] ,
does not follow any of the motifs described heretofore .
the unit cell consists of a @xmath72 square with motif of 5 squares attached to its side .
this 5-square pattern is also the best packing of 5 squares in a square @xcite .
a simple calculation yields the density , @xmath73 .
this packing has one square per unit cell tilted at 45@xmath74 relative to all other squares .
this is the only example that we found for which not all of the squares in the motif are oriented in the same way .
it was also the most difficult configuration for our simulated annealing algorithm to find .
figure [ fig : n21 ] shows a typical simulation result , which clearly has not yet fully converged . to summarize : the perfect square , sum of two squares and gapped bricklayer configurations cover most of the case we have found for @xmath7 .
the table gives densest packing configurations ( if @xmath75 ) and conjectured densest packing configurations ( if @xmath76 ) , for each value of @xmath0 less than 28 .
the column `` @xmath22 '' is the density of the configuration .
the comment column describes the type of lattice .
for example , @xmath77 indicates a perfect square and @xmath78 indicates a perfect square with one square missing .
similarly @xmath79 refers to the sum of two squares and `` gb '' stands for `` gapped bricklayer . ''
if a single number is in the comment column it refers to one of the special cases discussed above .
the four columns `` @xmath14 , @xmath15 , @xmath16 , @xmath17 '' are shown if the configuration of squares is itself a bravais lattice ; these integers are the coefficients of the lattice vectors of the torus , in terms of the lattice vectors of the squares as defined in eq .
( [ eqn : ana ] ) .
only those columns needed to specify the lattice are filled in .
[ table ] .exact and conjectured densest packed configurations of squares on a torus .
refer to the text for the meaning of the columns .
[ cols="^,^,^,^,^,>,^,^,^",options="header " , ]
for all @xmath7 squares on the flat torus , we searched for densest packings of @xmath0 squares on the torus via monte carlo simulations in the npt ensemble .
our approach was to employ a simulated annealing ( sa ) algorithm in which the system was taken from an initial , low - pressure , easy - to - equilibrate state to a final , high - pressure state , via an annealing schedule consisting of a series of steps in inverse pressure . between each pressure increase a metropolis algorithm appropriate to the hard square npt ensemble
was used to equilibrate the system .
although sa quickly falls out of equilibrium at higher pressures as the energy landscape becomes rough , it appears to be an an effective algorithm for finding ground states of the system .
the equilibration procedure we used in our simulated annealing algorithm was a metropolis procedure consisting of three types of monte carlo moves : translations and rotations of individual squares , and changes in the volume of the entire system . at each step of the equilibration procedure ,
a square is selected at random ; then , one of the three types of moves is selected at random , with probabilities .495 , .495 , and .01 , for translation , rotation , and volume change , respectively .
once a square and a move type is selected , the move is attempted . if the move results in any overlaps among the @xmath0 squares , the move is rejected .
if it does not , then for translations or rotations , the move is accepted ; for volume change @xmath80 , the move is accepted with probability @xmath81 $ ] . in practice , rather than changing the volume of the entire system , the periodic box in the simulations was kept at a constant size , and the sizes of the individual squares were all rescaled , in order to achieve the desired new volume .
the equilibration procedure consists of @xmath82 such monte carlo steps ; in our simluations @xmath82 was typically between 200 and 400 steps .
we repeated this simulation 1,000 times and reported the highest density found among these runs . in order to determine the highest - density packing for @xmath0 that did not correspond to a perfect square or a sum of squares
, more extensive runs were conducted in some cases , as long as 72 hours on a 2ghz processor . the fact that significantly different initial configurations as well as different initial random seeds generated the same final
, high - density configurations signalled that a good candidate for a densest packing of the system had been found .
for all simulations , the pressure was initially set to @xmath83 , and was increased via constant steps in inverse pressure until a maximum pressure of @xmath84 was reached .
( during subsequent explorations of the phase behavior of the system , this pressure was later deemed excessive , but nevertheless produced reasonable results for the purposes of determing the ground state of the system . )
all simulations were begun at an initial areal density @xmath22 of 0.1 , with a square array of unit squares .
before each equilibration procedure began , a trial run was conducted in which the maximum value of translations , rotations , and volume changes were independently optimized in order to achieve an acceptance ratio for each of @xmath85 . the majority of the computational work during the equilibration procedure consisted in checking for square overlaps .
for this , we relied on a fast algorithm for detecting polygon overlaps by alan murta @xcite , and an associated python wrapper by joerg rdler @xcite .
configurations were visualized using the vpython library @xcite .
in this paper , we have presented an analysis of the densest packing solutions for @xmath0 unit squares in the torus . for @xmath7 ,
the majority of these packings are bravais lattice solutions , manifested either as the `` sum of two square integers '' , or `` gapped bricklayer configurations '' described above .
a few were non - bravais lattice solutions , such as those @xmath71 and @xmath2 . in this section ,
we discuss the frequency and entropy of the various types of packings we found , and pose some questions for further study .
we showed in section [ sec : densityonepackings ] that density - one packings are only possible for those values of @xmath0 that are expressible as the sum of two squares .
though it appears that density - one packings are relatively common from the table , in fact it is known that the frequency of numbers that are equal to the sum of two square integers scales as @xmath86 for large @xmath0 @xcite .
thus the frequency of density - one packings also vanishes with increasing @xmath0 . despite the relative scarcity of density - one packings , we argue that the packing density approaches one as @xmath3 .
suppose @xmath87 is a sum of two square integers .
construct a new packing for @xmath88 by removing @xmath43 squares .
the resulting packing density gives a lower bound of @xmath89 .
given a number of squares , @xmath0 , however , determining @xmath43 requires knowledge of the nearest density - one packing larger than @xmath0 . for sufficiently large @xmath0
, we can estimate that the distance to the next density - one packing is of order @xmath90 larger than @xmath0 .
therefore , an estimated lower bound on the packing density of @xmath0 squares is @xmath91 for large @xmath0 , which yields the result that the density approaches one asymptotically . this argument does not take into account fluctuations in the spacing of sums of two squares , and it would be interesting to find a mathematically rigorous asymptotic lower bound on the packing density .
the main contributions to the entropy of the various bravais and non - bravais lattice configurations for @xmath7 are readily assessed . for a bravais lattice packing ,
there are no non - trivial continuous symmetries and thus there is no entropy .
however , one can construct density - one packings from the bravais lattice packings by shifting rows relative to one another .
for example , if @xmath0 is a perfect square then each row can be arbitrarily shifted and , ignoring overall shifts of the lattice , the entropy is proportional ( @xmath92 ) . note that either rows or columns may be shifted but not both for density - one packings .
more generally , rows are free to shift unless the contraints of periodicity forbid it .
periodicity forbids two rows from shifting relative to one another if some linear combination of torus lattice vectors connects the rows .
suppose that rows are aligned in the @xmath93-direction .
then , two rows are constrained to their relative positions in the bravais lattice configuration if there are integers @xmath94 and @xmath95 such that @xmath96 connects the two rows .
thus , rows separated by @xmath97 are locked ( see eq .
[ eqn : ana ] and fig .
[ fig : aligned ] ( b ) ) .
bezout s lemma @xcite states that this integer linear combination can be made equal to , but not smaller than , @xmath98 , where @xmath98 is the greatest common divisor of @xmath15 and @xmath17 ( assuming that both @xmath15 and @xmath17 are nonzero ) .
thus , the set of rows can be divided into @xmath98 groups , each of which can be arbitrarily shifted , and the entropy is proportional to @xmath99 .
note that if @xmath100 and @xmath101 are mutually prime , all rows are locked and this contribution to the entropy of the configuration vanishes .
there are two other sources of entropy for packing related to bravais lattice packings .
the gapped bricklayer configurations ( section [ sec : bricklayer ] ) allow squares within a row to shift perpendicular to the row axis ( see figures [ fig : gb ] and [ fig : n22 ] ) , and this `` poor workmanship '' contribution to the entropy will be roughly proportional to the free volume of the configuration .
the density - one configurations with vacancies ( @xmath102 ) , discussed in section [ sec : vacancies ] , also possess finite entropy , since the hole(s ) created by the @xmath43 missing squares may be moved throughout the lattice , or split along a row ; and for @xmath103 , holes can appear in different rows .
in contrast , the unusual , non - bravais lattice packings of the @xmath62 and @xmath2 exhibit no entropy .
the above entropies have implications for which configurations are most likely to be seen at finite pressure .
for example , @xmath104 admits of two classes of configuration : a density - one packing with rows aligned along the torus ( @xmath105 ) or with rows oriented at an angle with respect to the torus ( @xmath106 ) .
as we have seen , the @xmath105 has four rows that are free to slide while @xmath107 is a locked configuration with no entropy ; this implies that the @xmath105 packings will be much more likely to appear at finite pressure . unlike squares packed into a square boundary ,
squares packed on a torus maintain rotational invariance in the thermodynamic limit .
this can be seen as follows : one can see in figure [ fig : bravais ] that any @xmath4 packing ( which are , as seen above , the only possible packings with density one ) will orient the square lattice at an angle of @xmath108 relative to the torus lattice vectors . to take the thermodynamic limit with a particular square lattice orientation @xmath109 with respect to the underlying torus , it is sufficient to choose a particular subsequence of integers @xmath110 such that @xmath111 .
the thermodynamic limit of density - one packings on the torus thus preserves rotational symmetry .
our study of the densest packings of @xmath0 unit squares in a torus has yielded definitive results for cases in which @xmath0 is the sum of two square integers or is a perfect square , and strong conjectures for other values of @xmath112 .
this work raises many interesting questions : how common are densest packings that have squares with different orientations , such as occurs for @xmath71 ?
which motifs , if any , dominate for large @xmath0 ?
are the @xmath113 motifs of @xmath62 and @xmath2 exhibited for other @xmath0 ?
what role do these dense packing configurations play in the thermodynamic phases observed in finite - temperature simulations and experiments with square colloids ?
we would like to thank james hanna and narayanan menon for useful discussions of some of the issues addressed in this paper . j.m .
acknowledges support from nsf grant dmr-0907235 .
c.d.s . acknowledges support from nsf grant dmr-0846582 , and was partially supported by the nsf - funded center for hierarchical manufacturing , cmmi-1025020 .
d.b . is grateful for use of the hoffman2 computing cluster at ucla . | the densest packings of @xmath0 unit squares in a torus are studied using analytical methods as well as simulated annealing .
a rich array of dense packing solutions are found : density - one packings when @xmath0 is the sum of two square integers ; a family of `` gapped bricklayer '' bravais lattice solutions with density @xmath1 ; and some surprising non - bravais lattice configurations , including lattices of holes as well as a configuration for @xmath2 in which not all squares share the same orientation .
the entropy of some of these configurations and the frequency and orientation of density - one solutions as @xmath3 are discussed . |
in a recent paper , @xcite we showed that the traditional concept of the uniform electron gas ( ueg ) , _ i.e. _ a homogeneous system of finite density , consisting of an infinite number of electrons in an infinite volume , @xcite is inadequate to model the uegs that arise in finite systems .
accordingly , we proposed to embark on a comprehensive study of quasi - exact properties of finite - size uegs , in order eventually to create improved approximations in density - functional theory .
@xcite in an earlier paper , @xcite we introduced an alternative paradigm , in which @xmath0 electrons are confined to a @xmath11-sphere ( with @xmath12 ) , that is , the surface of a ( @xmath13)-dimensional ball .
these systems possess uniform densities , even for finite @xmath0 and , because all points on a @xmath11-sphere are equivalent , their mathematical analysis is relatively straightforward .
@xcite in the present paper , we study the one - dimensional ( @xmath14 ) version of model , in which @xmath0 electrons are confined to a ring of radius @xmath15 .
the electron density of this @xmath0-electron ueg , which we will call @xmath0-ringium , is @xmath16 where @xmath17 is the seitz radius . in most previous work on the one - dimensional ( 1d ) ueg
, the true coulomb potential @xmath18 has been avoided because of the intractability of its fourier transform . instead , most workers have softened the potential , either by adding a transverse harmonic component @xcite or by using a potential of the form @xmath19 . in the latter case ,
the @xmath20 parameter eliminates the singularity at @xmath21 while retaining the long - range coulomb tail . @xcite however , the introduction of a parameter @xmath22 is undesirable , for it modifies the physics of the system in the high - density regime where neighboring electrons repel far too weakly .
it is also unnecessary , because the true coulomb potential is so repulsive that it causes the wave function to vanish when any two electrons touch , thereby removing the possibility of an energy divergence .
@xcite this nodal behavior leads to the 1d bose - fermi mapping @xcite which states that the ground state wave function of the bosonic ( b ) and fermionic ( f ) states are related by @xmath23 , where @xmath24 are the one - particle coordinates . in case of bosons , the divergence of the coulomb potential has the effect of mimicking the pauli principle which prohibits two fermions from touching .
this implies that , for 1d systems , the bosonic and fermionic ground states are degenerate and the system is `` spin - blind '' . consequently , the paramagnetic and ferromagnetic states are degenerate and we will consider only the latter .
@xcite the electrons - on - a - ring paradigm has been intensively studied as a model for quantum rings ( qrs ) , which are tiny , self - organised , ring - shaped semiconductors @xcite characterised by three parameters : radius ( @xmath15 ) , width ( @xmath25 ) and electron number ( @xmath0 ) .
modern microfabrication technology has yielded ingaas and gaalas / gaas qrs that bind only a few electrons , @xcite in contrast with the mesoscopic rings on gaas which hold much larger numbers of electrons .
@xcite these low - dimensional systems are the subject of considerable scientific interest and have been intensively studied , both experimentally @xcite and theoretically , @xcite mainly because of the observation of aharonov - bohm oscillations .
@xcite as a first approximation , qrs can be modelled by electrons confined to a perfect ring ( _ i.e. _ @xmath26 ) .
in a recent paper , @xcite we considered a pair of electrons ( _ i.e. _ @xmath27 ) on such a ring and discovered that their schrdinger equation can be solved exactly , provided that the radius takes one of an infinite number of special values . some of the solutions exhibit the berry phase phenomenon , _
i.e. _ if one of the electrons moves once around the ring and returns to its starting point , the wave function of the system changes sign .
qrs are among the simplest systems with this peculiar property . in section [ sec : pert ]
, we first use rayleigh - schrdinger perturbation theory to investigate the energy in the high - density regime @xcite and then strong - coupling perturbation theory to study the low - density regime , where the electrons form a wigner crystal .
@xcite in section [ sec : ec ] , we use explicitly correlated ( ec ) methods to determine the energy of @xmath0-ringium for @xmath28 .
these methods are accurate for very small @xmath0 but their cost grows very rapidly with @xmath0 . in section
[ sec : qmc ] , we turn to quantum monte carlo ( qmc ) approaches for studying @xmath0-ringium up to @xmath29 over a range of densities .
these methods provide a different approach to the many - body problem : variational monte carlo ( vmc ) @xcite and diffusion monte carlo ( dmc ) @xcite methods can be used to treat systems in one and higher dimensions at a computational cost that grows relatively slowly with @xmath0 ( at least , when @xmath0 is not too large .
@xcite ) we frame our discussion in terms of _ reduced _ energy @xmath30 , _ i.e. _ energy per electron , so that we can pass smoothly from finite to infinite @xmath0 .
one of the key goals of the paper is to develop an understanding of the correlation energy , which is defined as the difference @xmath31 between the exact and hartree - fock ( hf ) energies .
atomic units are used throughout , but we report total energies in hartrees ( @xmath32 ) and correlation energies in millihartrees ( @xmath33 ) .
the hamiltonian of the system is @xmath34 where @xmath35 is the angle of electron @xmath36 around the ring center , and @xmath37 is the across - the - ring distance between electrons @xmath36 and @xmath38 . in the high - density ( _ i.e. _ small @xmath2 ) regime , the kinetic energy is dominant and it is natural to define a zeroth - order hamiltonian @xmath39 and a perturbation @xmath40 the non - interacting orbitals and orbital energies are @xmath41 where @xmath42 a slater determinant @xmath43 of any @xmath0 of these orbitals has an energy @xmath44 and is an antisymmetric eigenfunction of @xmath45 . in the lowest energy ( aufbau )
determinant @xmath46 , we occupy the orbitals with @xmath47 following the approach of mitas , @xcite one discovers the remarkable result @xmath48 where @xmath49 is a signed interelectronic distance .
it follows immediately that @xmath46 has a node whenever @xmath50 and , therefore , possesses the same nodes as the exact wave function .
this will have important ramifications in section [ sec : qmc ] .
rayleigh - schrdinger theory yields the perturbation expansion for the reduced energy @xmath51 where the high - density coefficients @xmath52 are found by setting @xmath53 and evaluating @xmath54 to evaluate the coefficients @xmath52 with @xmath55 , one requires the `` double - bar '' integrals @xmath56 } { r_{12 } } d\theta_1 d\theta_2.\ ] ] by elementary integration , one can show that @xmath57 where @xmath58,\ ] ] and @xmath59 is the digamma function .
@xcite .[tab : eps ] high - density coefficients for @xmath0-ringium .
( @xmath60 is apry s constant .
@xcite ) [ cols="^,^,^,^,^ " , ]
we have studied @xmath0-ringium using explicitly correlated and quantum monte carlo methods . using hylleraas wave functions ,
we have obtained the near - exact ground - state energy of the @xmath61 and @xmath62 systems for various values of the seitz radius @xmath2 . for @xmath63 ,
we have performed exact - node dmc calculations to find the exact ground - state energies , with statistical errors in the @xmath64 range .
we have shown that the reduced correlation energy of @xmath0-ringium is @xmath65 for high densities , and @xmath66 for low densities .
expressions for the coefficients are given in eqs .
, , , and . in the thermodynamic limit , we have found that @xmath67 and shown that the ringium and jellium models are equivalent in the thermodynamic limit .
this provides a detailed picture of the energy of this new model over a wide range of @xmath0 and @xmath2 values and we believe that the correlation energies in table [ tab : ec ] are the most accurate yet reported for @xmath0-ringium .
these systems are distinct uniform electron gases @xcite and can be used to design a new correlation functional for 1d systems .
we will report such a functional in a forthcoming paper .
@xcite the authors thank neil drummond and shiwei zhang for helpful discussions , the nci national facility for a generous grant of supercomputer time .
pmwg thanks the australian research council ( grants dp0984806 , dp1094170 , and dp120104740 ) for funding .
pfl thanks the australian research council for a discovery early career researcher award ( grant de130101441 )
for the ground state , the hamiltonian can be recast as @xmath68 + \sum_{\substack{i \neq j\\ i \neq k}}^{n } \sum_{j < k}^{n}\frac{r_{ik}^2 + r_{jk}^2 - r_{ij}^2}{2\,r_{ik}r_{jk } } \frac{r_{ij}^2 + r_{jk}^2 - r_{ik}^2}{2\,r_{ij}r_{jk } } \frac{\partial^2}{\partial r_{ij } \partial r_{ik}}.\ ] ] the first term in contains the two - body parts of the hamiltonian while the second includes coupling between electron pairs .
the @xmath0-electron overlap integrals needed in calculations on @xmath0-ringium can be systematically constructed using the unit - ring fourier resolution @xmath69 where @xmath70 is a signed binomial coefficient .
. is valid for @xmath71 and terminates if @xmath72 is an even integer . resolving each integrand factor , swapping the order of summation and integration , performing the integrations and resumming
, often leads to beautiful expressions .
for example , the cyclic @xmath0-electron integral yields @xmath73 which can be written as a @xmath74 hypergeometric function of unit argument .
@xcite in some cases , the sums can be found in closed form , for example , @xmath75 but this is not possible in general .
@xcite ( eq . uses the result of problem 62 of knuth s book .
@xcite ) however , sums such as converge rapidly and are numerically satisfactory .
76ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ]
+ 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty @noop * * , ( ) @noop _ _ ( , , ) @noop * * , ( ) @noop * * , ( ) @noop _ _
( , ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) `` , '' ( , ) pp . @noop * * , ( ) `` , '' ( , ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) , , , and , eds . ,
@noop _ _ ( , , ) @noop * * , ( ) link:\doibase 10.1103/physrevb.15.2819 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop , @noop _ _ ( , , ) @noop _ _ , vol . | we introduce a new paradigm for one - dimensional uniform electron gases ( uegs ) . in this model , @xmath0
electrons are confined to a ring and interact via a bare coulomb operator .
we use rayleigh - schrdinger perturbation theory to show that , in the high - density regime , the ground - state reduced ( _ i.e. _ per electron ) energy can be expanded as @xmath1 , where @xmath2 is the seitz radius .
we use strong - coupling perturbation theory and show that , in the low - density regime , the reduced energy can be expanded as @xmath3 .
we report explicit expressions for @xmath4 , @xmath5 , @xmath6 , @xmath7 , @xmath8 and @xmath9 and derive the thermodynamic ( large-@xmath0 ) limits of each of these .
finally , we perform numerical studies of uegs with @xmath10 , using hylleraas - type and quantum monte carlo methods , and combine these with the perturbative results to obtain a picture of the behavior of the new model over the full range of @xmath0 and @xmath2 values . |
there are four ways by which extrasolar planets have been detected .
the first method to unambiguously detect an extrasolar planet was pulsar timing , which relies on detecting periodic variations in the timing of the received radio signal that occur as the pulsar orbits about the system s barycenter .
the first system of three planets was found around psr b1257 + 12 in 1992 @xcite , followed by a single planet around psr b1620 - 26 @xcite .
although rare , the pulsar planets are some of the lowest mass extrasolar planets known : psr b1257 + 12a is about twice the mass of the moon .
the second way to find extrasolar planets is through radial velocities ( rv ) , which uses the doppler shift of observed stellar spectra to look for periodic variations in the target star s radial velocity . by estimating the mass of primary star , the observed radial velocity curve and velocity semi - amplitude
can then be used to directly calculate the inclination - dependent mass ( @xmath0 ) of the companion object . to date
, rv surveys have detected more than 230 planets around other stars , making it the most successful method of extrasolar planet detection . while the large number of detected systems having an unseen companion with a mass on the order of @xmath1 statistically ensures that the majority of these are planetary bodies , the rv surveys are unable to provide more information than the minimum masses , periods , eccentricities , and the semi - major axes of the planets .
rv surveys also have a limited ability to detect planets much smaller than a few earth masses .
the state of the art in rv surveys is the high accuracy radial velocity planet searcher ( harps ) spectrometer at the la silla observatory in chile , which is capable of radial velocity measurements with precisions better than @xmath2 for extended periods of time @xcite .
harps is therefore able to detect planets with masses on the order of @xmath3 to @xmath4 in relatively short period orbits .
unfortunately , planets closer to an earth mass will be increasingly difficult to detect since intrinsic stellar variability , in the form of acoustic oscillation modes and granulations on the photosphere , makes more precise spectroscopic radial velocity measurements harder to acquire .
however , it may be possible to surmount this obstacle ( as in the case of @xcite ) through the selection of stars with `` quiet '' photospheres and long integration times which serve to average out the stellar variability .
gravitational microlensing is another technique for detecting extrasolar planets .
microlensing of a star occurs when a a star passes near the line of sight of the observer to another background star .
the gravity of the foreground star acts as a lens on the light emitted by the background star , which causes the star in the background to become momentarily brighter as more light is directed towards the observer .
planetary companions to the lens star can further magnify the background star , and create short - term perturbations to the microlensing light curve . to date , six planets have been detected using microlensing @xcite .
unfortunately , the one - shot nature of microlensing observations means that information about systems discovered this way is generally sparser than that available for rv systems .
therefore , microlensing is most useful in determining the general statistical properties of extrasolar planets , such as their frequency and distribution , and not the detailed properties of the planetary systems .
planetary transits are a fourth method by which extrasolar planets have been discovered , and the one that provides the most complete set of information about the planetary system
. only planets with very specific orbital characteristics have a transit visible from earth , because the orbital plane has to be aligned to within a few degrees of the line of sight . therefore transiting planets are rare .
nevertheless , a transiting extrasolar planet offers the opportunity to determine the mass of the planet ( when combined with rv measurements ) since the inclination is now measurable , as well as the planetary radius , the density , the composition of the planetary atmosphere , the thermal emission from the planet , and many other properties ( see @xcite for a review ) . additionally , and unlike rv surveys ,
transiting planets should be readily detectable down to @xmath5 and beyond , even for relatively long periods .
having accurate predictions of the number of detectable transiting planets is immediately important for the evaluation and design of current and future transit surveys . for the current surveys ,
predictions allow the operators to judge how efficient are their data - reduction and transit detection algorithms .
future surveys can use the general prediction method that we describe here to optimize their observing set - ups and strategies .
more generally , such predictions allow us to test different statistical models of extrasolar planet distributions .
specifically , as more transiting planets are discovered and characterized , predictions relying on incorrect statistics of extrasolar planet properties will increasingly diverge from the observed set of distributions .
using straightforward estimates it appears that observing a planetary transit should not be too difficult , presuming that one observes a sufficient number of stars with the requisite precision during a given photometric survey .
specifically , if we assume that the probability of a short - period giant planet ( as an example ) transiting the disk of its parent star is 10% , and take the results of rv surveys which indicate the frequency of such planets is about 1% @xcite , together with the assumption that typical transit depths are also about 1% , the number of detections should be @xmath6 , where @xmath7 is the number of surveyed stars with a photometric precision better than 1% .
unfortunately , this simple and appealing calculation fails . using this estimate
, we would expect that the tres survey , which has examined approximately 35,000 stars with better than 1% precision , to have discovered 35 transiting short period planets .
but , at the date of this writing , they have found four .
indeed , overall only 51 transiting planets have been found at this time by photometric surveys specifically designed to find planets around bright stars .
this is almost one hundred times less than what was originally predicted by somewhat more sophisticated estimates @xcite .
clearly then , there is something amiss with this method of estimating transiting planet detections .
several other authors have developed more complex models to predict the expected yields of transit surveys .
@xcite examined the potential of all - sky surveys , which was expanded upon and generalized for photometric searches in clusters @xcite .
@xcite and @xcite tested whether the ogle planet detections are statistically consistent with radial velocity planet distributions .
@xcite was the first to make published estimates of the rate of false positives in transit surveys , and @xcite model transit detections to estimate and compare the potential of several ground- and space - based surveys . as has been recognized by these and other authors , there are four primary reasons why the simple way outlined above of estimating surveys yields fails .
first , the frequency of planets in close orbits about their parent stars ( the planets most likely to show transits ) is likely lower than rv surveys would indicate .
recent examinations of the results from the ogle - iii field by @xcite and @xcite indicate that the frequency of short - period jovian - worlds is on the order of @xmath8 , not @xmath9 as is often assumed by extrapolating from rv surveys @xcite .
@xcite point out that most spectroscopic planet searches are usually magnitude limited , which biases the surveys toward more metal - rich stars , which are brighter at fixed color .
these high metallicity stars are expected to have more planets than solar - metallicity stars @xcite .
second , a substantial fraction of the stars within a survey field that show better than 1% photometric precision are either giants or early main - sequence stars that are too large to enable detectable transit dips from a jupiter - sized planet @xcite .
third , robust transit detections usually require more than one transit in the data .
this fact , coupled with the small fraction of the orbit a planet actually spends in transit , and the typical observing losses at single - site locations due to factors such as weather , create low window probabilities for the typical transit survey in the majority of orbital period ranges of interest @xcite .
lastly , requiring better than 1% photometric precision in the data is not a sufficient condition for the successful detection of transits : identifiable transits need to surpass some kind of a detection threshold , such as a signal - to - noise ratio ( s / n ) threshold .
the s / n of the transit signal depends on several factors in addition to the photometric precision of the data , such as the depth of the transit and the number of data points taken during the transit event .
additionally , ground - based photometry typically exhibits substantial auto - correlation in the time series data points , on the timescales of the transits themselves .
this additional red noise , which can come from a number of environmental and instrumental sources , substantially reduces the statistical power of the data @xcite . in this paper
we describe a method to statistically calculate the number of planets that a given transiting planet survey should be able to detect .
we have done so in a way that maximizes the flexibility of our approach , so that it may be applied in as many circumstances as possible .
therefore , we have allowed for survey - specific quantities such as the observation bandpass to be defined arbitrarily , and have also allowed for our astrophysical assumptions to be easily altered .
we account for factors such as the low frequency of jovian planets , the increased difficulty of detecting transits at high impact parameters , as well as variations in stellar density due to galactic structure and extinction due to interstellar dust .
we also use the present - day mass function to realistically account for the number of main - sequence stars that will be in a given field of view .
we do not treat , or simulate , the distribution and effects of giant stars .
we approximate the detectability of a planetary transit using a s / n formulation .
this differs from the previously published work in the field by virtue of the generalized nature of our approach .
other authors , such as @xcite , @xcite , @xcite , and @xcite have restricted themselves to modeling specific transit surveys or types of transit surveys .
our approach , which allows for variations such as arbitrary telescope parameters , observing bandpasses , cadences , and fields of view , is not restricted to any one type of transit survey .
indeed , we offer examples of how our method may be applied to the four most frequently used modes for transit searches .
furthermore , we have conducted an extensive literature review of basic astrophysical relations , such as the present day mass - function ( pdmf ) , galactic structure , and the main sequence mass - luminosity relation , and included these in our predictions .
we do not use more complicated ( but presumably more accurate ) galactic models like the besancon model @xcite because these models are less flexible and considerably more time consuming to utilize .
we first describe the mathematical formalism with which we have chosen to address this problem , and then move on to discuss our specific assumptions . as examples
, we then offer our predictions for several different types of photometric surveys : point - and - stare , drift - scanning , space - based , and all - sky . for each ,
we choose a representative survey for that particular mode , and compare our predictions to actual survey results , or to predictions elsewhere in the literature .
we also examine the possible effects of red noise on our predictions , and conclude by discussing what our predictions imply for the general state of photometric transiting planet surveys .
in the most general sense , we can describe the average number of planets that a transit survey should detect as the probability of detecting a transit multiplied by the local stellar mass function , integrated over mass , distance , and the size of the observed field ( described by the galactic coordinates @xmath10 : @xmath11 where @xmath12 is the probability that a given star of mass @xmath13 and distance @xmath14 orbited by a planet with radius @xmath15 and period @xmath16 will present a detectable transit to the observing set - up .
@xmath17 is the probability that a star will possess a planet of radius @xmath15 and period @xmath16 .
@xmath18 is the present day mass function in the local solar neighborhood , and @xmath19 is the local stellar density for the three - dimensional position defined by ( @xmath20 ) .
we use @xmath21 instead of the usual volume element for spherical coordinates , @xmath22 , because @xmath23 is defined opposite to @xmath24 : @xmath25 occurs at the pole . first , we will integrate over both planetary radius and orbital period @xmath26 to do this , we must first specify the detection probability @xmath27 , as well as the distribution of planet radius and periods @xmath28 , which we treat in section 2.2 . the probability
that a planet orbiting a given star shows a detectable transit may be broken up into two separate probabilities : the chance that a planet with radius @xmath15 and period @xmath16 around a star of mass @xmath13 at a distance @xmath14 will transit with a sufficient signal - to - noise ratio to be detected , and the window probability that a transit will be visible for the particular observing set - up : @xmath29 to calculate the probability that a planetary system will show detectable transits , we first select the appropriate statistical test to determine whether or not our notional data shows a discernible transit . in the world of transiting planet searches ,
many different methods are used either individually or together to test the statistical significance of a possible transit and alert the researcher to its existence .
for example , the hungarian - made automated telescope network ( hatnet ) uses a combination of the dip significance parameter ( dsp ) and a box - least squares ( bls ) power spectrum ( amongst others ) to test its photometric data . in general
though , whatever test one uses , it is generally the s / n of a transit shape in the observed data is what determines whether or not that transit will be identified as an event .
as the square of the signal - to - noise can be approximated by the @xmath31 test , we therefore chose to use the @xmath31 value to determine if a given transit will be detectable for a set of given observation parameters .
for a specific transit , the @xmath31 statistic , assuming uncorrelated white noise , is given by @xmath32 where @xmath33 is the number of data points observed in a transit of fractional depth @xmath34 and with fractional standard deviation @xmath35 , assuming a boxcar transit shape with no ingress or egress . to be detected , the transit must have a @xmath31 higher than a certain minimum @xmath36 @xmath37 the fractional depth of the transit is given by the square of the planet - to - star radius ratio @xmath38 while this formulation ignores the contributions of stellar limb - darkening to the depth of the transit , @xcite demonstrate that , to a first order approximation , transit detection rates are not affected by limb - darkening .
briefly , they point out that the effects of limb - darkening on the transit lightcurve will be offset by the effects of the planet s ingress and egress from the stellar disk .
therefore , we have not included stellar limb - darkening terms in our calculations . to determine @xmath35 , we assume poisson statistics , such that for a given exposure with an observed number of @xmath39 source photons from the target star and @xmath40 background photons , @xmath41 where @xmath42 accounts for scintillation as per @xcite , and @xmath39 is defined as @xmath43 here @xmath44 is the efficiency of the overall observing set - up , and can take values from 0 to 1 .
it accounts for photon losses in places such as the filter , the mirrors of the telescope , the ccds , and passage through the atmosphere .
@xmath45 is the source photon flux in the observed bandpass in units of @xmath46 .
@xmath47 is the exposure time of each observation , and @xmath48 is simply the light collecting area of the telescope .
@xmath45 , the observed photon flux from the source in a certain bandpass , can be defined as a function of the photon luminosity of the source , the distance to the source , and the effects of extinction due to interstellar dust : @xmath49 where @xmath50 accounts for extinction , and will be treated later .
it has the effect of decreasing the number of observed photons from more distant stars , particularly in the bluer bands .
@xmath51 , the photon luminosity of the source star , we set as a function of the stellar mass @xmath13 and the observational bandpass @xmath52 .
we derive this quantity in appendix a , and only note here that we approximate the source stars as blackbodies . while this is a simplification , we show in appendix a that it is sufficient for our purposes .
using @xmath51 in equation ( [ eq:80 ] ) yields the observed photon flux at a distance @xmath14 from that same star .
we thus now have an expression for @xmath39 as a function of the stellar mass , distance , and bandpass choice , that we can use in our calculation of @xmath35 .
we now only need @xmath40 , the number of background photons observed .
similar to equation ( [ eq:70 ] ) , @xmath40 is defined as @xmath53 where @xmath54 is the photon surface brightness of the sky at the particular wavelength @xmath52 ( or , in our case , for a specific band ) , and @xmath55 is the effective area of the seeing disk .
assuming that the point - spread function ( psf ) is a gaussian with a full - width half - max ( fwhm ) of @xmath56 arcseconds , @xmath55 is defined by @xmath57 we can now place both of the expressions for @xmath39 and @xmath40 back into equation ( [ eq:60 ] ) to arrive at a determination of @xmath35 , the photometric uncertainty in the data points .
in certain cases , @xmath39 and @xmath40 will be so great that the ccd detectors of the telescope will be saturated by the number of photons received from a star during the exposure time . in this case ,
further measurements become useless .
mathematically , we may describe saturation as occurring when the number of photons collected by a single ccd pixel during an exposure exceeds the full - well depth ( fwd ) of that pixel : @xmath58 @xmath59 is related to @xmath39 and @xmath40 as @xmath60\right)\nonumber \\ & & + \frac{n_b \theta_{pix}^2}{\omega } , \end{aligned}\ ] ] where @xmath61 is the angular size of an individual pixel .
the only part of equation ( [ eq:30 ] ) that remains to be defined is the number of data points observed while the system is in transit , @xmath33 . assuming a circular orbit
, we can say that the number of points seen in transit is directly proportional to the fraction of time that the system spends in eclipse , and given a total number of observed data points @xmath62 , @xmath33 is thus @xmath63 where @xmath23 is the impact parameter of the transit in units of the stellar radius , @xmath64 , @xmath65 is the orbital inclination , and we have assumed that @xmath66 .
we note that this does not hold for correlated noise , see @xcite .
taking equation [ eq:160 ] and putting it back into our original formulation for @xmath31 gives @xmath67 we define a new variable , @xmath68 , as the value of equation ( [ eq:170 ] ) when @xmath69 , corresponding to the @xmath31 of an equatorial transit about a given star .
this gives us a new form that makes it easier to determine when a system will have a transit geometry sufficient to be detected by the given observing set - up : @xmath70 rearranging , we find that we are able to detect transits up to a @xmath71 of @xmath72 therefore , @xmath71 is a function of @xmath13 , @xmath14 , and @xmath16 as a result of the relations contained within the @xmath68 term . finally , we can now deduce an explicit expression for @xmath30 .
given that detectable transits can only occur between @xmath73 , we may simply integrate the normalized probability density function of allowable impact parameters from 0 to @xmath71 to find a value for @xmath30 . given that @xmath74 , the maximum value of the impact parameter for a circular orbit ( different from the maximum at which we will detect transits ) is @xmath75 .
the normalization for the impact parameter s probability density function is therefore @xmath76 since impact parameters are uniformly distributed is uniform , which follows from the assumption that planetary systems are randomly oriented .
the distribution of @xmath77 may be derived by considering the normalized angular momentum vector of the planetary orbit , which is free to point in any direction , and thus has a parameter space that forms an imaginary sphere around the parent star .
particular values of @xmath77 are `` hoops '' along the surface of this sphere , and the ratio of their area to that of the total sphere is the probability that @xmath77 will take that value . ] , this gives the solution @xmath78 .
it should be noted that this form of @xmath79 is only valid in the case of circular orbits . for
the period ranges probed so far by transit surveys , this is not a very restrictive assumption , since only @xmath80 of the known transiting planets have eccentricities inconsistent with zero . at longer periods , however , the effects of eccentricity must be considered @xcite .
if we now integrate this over the range of impact parameters that yield detectable transits , we find @xmath81 the window probability ( or window function ) for a transit survey describes the probability that a transiting planet with a certain period will be visible to the survey , given the observational cadence of the survey ( i.e. - x hours of observing per night for y nights ) . while many numerical routines exist that can exactly calculate the window probability for a given observing set - up , for our purposes calculating an exact numerical answer is generally too time intensive to be practical .
-0.0 in -0.0 in we therefore developed an analytical framework that allows for the approximate calculation of a window probability .
this has the virtue of requiring much less computing time .
this is gained at the cost of a small loss of accuracy .
we give a full description of our analytic window probability in appendix b. briefly , we assume randomly sampled observations , which leads to simple expressions for the average window probability .
this implicitly ignores the effects of aliasing at integer and half - integer day periods .
thus we are not able to address issues related to non - uniform or biased _ a posteriori _ distributions of planet parameters resulting from the effect of aliasing ( e.g. @xcite ) .
nevertheless , the analytic solution does follow the `` outline '' of the exact calculations , and closely tracks the moving average value of the exact calculations sufficiently accurately for our purposes , as is shown in figure 1 .
@xmath83 denotes the probability that a given star has a planet with radius @xmath15 and orbital period @xmath16 . in the example surveys that we consider later
, we chose to focus on the detection of short period very hot- and hot - jupiters ( vhjs and hjs ) with orbital periods of 1 to 5 days , as this is the regime being probed by these surveys .
the exception of this focus is the kepler survey , where we also examine the detectability of transiting planets over a range of radii and orbital periods . in both cases , we consider a fixed planet radius , so that @xmath84 where @xmath85 is the correct normalization for a given period range , @xmath86 is the distribution within this range , and @xmath87 is the planet radius .
for vhjs and hjs we use @xmath88 , the average radius of the transiting planets found to date . in general , we have left the exact distribution of @xmath86 to be assigned as is deemed appropriate . for the vhj and hj period distributions described above ,
we adopt a distribution that is uniform in period , @xmath89 . the frequency of hot jupiters ( @xmath90 , @xmath91 ) is commonly taken to be @xmath92 from the results of the rv planet surveys ( e.g. , @xcite ) .
however , adopting frequencies from rv surveys can lead to biased estimates of the expected yields from transit surveys , because rv surveys constrain the frequency of planets as a function of mass , rather than radius .
furthermore , as argued by @xcite , most rv surveys are more metallicity biased than transit surveys , and therefore the planet frequency found by rv surveys will be correspondingly higher , since planet occurrence is an increasing function of metallicity @xcite .
rv surveys are metallicity biased because the target samples for these surveys are magnitude - limited . since metal rich stars are brighter at fixed color , they are overrepresented in magnitude - limited samples with respect to a volume - limited sample @xcite .
indeed , @xcite note that because they use a magnitude - limited target list , their target stars have a definite metallicity bias .
transit surveys , on the other hand , are typically signal - to - noise ratio limited . at fixed color ,
metal - rich stars are both brighter and bigger .
these two effects on the signal - to - noise ratio roughly cancel , and as a result transit surveys are very weakly biased with respect to metallicity ( see section 8.1 of @xcite for a quantitative discussion . ) indeed , there is evidence for the effects of this metallicity bias in the comparison of various surveys . from the california and carnegie keck magnitude - limited and metallicity - biased @xcite survey sample ,
@xcite find a frequency of @xmath93 for planets with @xmath94 and @xmath95 . on the other hand , for the volume - limited @xcite coralie sample , @xcite report a frequency of @xmath96 for @xmath95 and @xmath97 @xmath98 . assuming that @xmath99 of these planets have @xmath90
, this corresponds to a frequency of 0.6% for @xmath90 and @xmath95 , a factor of @xmath100 lower than that found for the keck rv survey .
@xcite examined the ogle - iii transit survey , and found a frequency of @xmath101 for @xmath90 , which was subsequently confirmed and strengthened by @xcite .
thus , as would be expected based on the arguments above , the frequencies inferred from transit surveys and volume - limited rv surveys are similar ( and entirely consistent considering poisson uncertainties and uncertainties in the mass - radius relation of giant planets ) , whereas they are substantially lower than the frequencies inferred from magnitude - limited ( and metallicity - biased ) rv surveys . for the vhj and hj predictions that we later make using tres , xo , and kepler
, we will adopt the normalizations from @xcite of @xmath102 for vhjs with orbital periods uniformly distributed between 1 - 3 days , and @xmath103 for hjs uniformly distributed between 3 - 5 day periods . by using statistics from ogle - iii , instead of the rv statistics , we are avoiding the difficulties discussed above , and furthermore any unaccounted - for biases inherent in photometric surveys will be preserved , and therefore any systematic effects unknown to us will still be accurately captured in the final predictions . in other words , it is an `` apples - to - apples '' comparison , instead of `` apples - to - oranges '' using rv data .
the next step is to integrate equation ( [ eq:10 ] ) over mass : @xmath104 having previously derived an expression for @xmath27 , we need only determine the appropriate formulation for the pdmf , @xmath105 . using the results of @xcite , who used the hipparcos data sets along with their own data from the palomar / michigan state university survey to create a volume - limited survey out to 25 pc , we adopt @xmath106 where @xmath107 is the appropriate normalization constant .
a fuller discussion for our decision to use this particular form for the pdmf is in appendix c. adopting a stellar mass density in the solar neighborhood of @xmath108 @xcite , we find a normalization of @xmath109 .
while our normalization of the mass function properly describes the density of stars in the solar neighborhood , beyond about 100 pc we must account for variations in stellar density due to galactic structure .
additionally , as mentioned previously , we must also take into consideration the effects of extinction due to interstellar dust .
this is the final stage of integration that needs to be done on equation ( [ eq:10 ] ) : @xmath110 we adopt the bahcall & soneira model for the galactic thin disk @xcite .
the model treats stellar density in the thin disk as a double exponential function involving both the distance from the galactic center and height above the galactic plane .
specifically , the mass density relative to the local density is given by @xmath111,\ ] ] where @xmath112 kpc is the distance from the galactic center .
@xmath113 is the scale length , and @xmath114 is the scale height .
we adopt a value of @xmath115 for the scale length of the disk . for the vertical scale height ,
we use the following relation , which is a slightly modified form of that originally proposed by @xcite : @xmath116 we calculate the absolute v - band magnitudes of stars of a given mass using the blackbody assumptions , and the mass - radius and mass - luminosity relations described in appendix a. we defer the complete discussion of our reasoning behind our choices for @xmath113 and @xmath114 to appendix d. the exact values for these parameters are not well constrained in the literature , and we show the effects of using different values in the appendix .
we note that within the range of values found in the literature , the final prediction results may differ by as much as 20% .
we only treat the thin disk in our modeling because we decided that an expanded analysis that also looked at stars in the thick disk , halo , and bulge would not add a significant number of detections to our final result .
both the thick disk and the halo have much lower volume densities than the thin disk at distances typically surveyed by transit searches . in the same vein ,
the bulge is too far away .
these populations are also typically older than the thin disk , and so have a fainter turn - off point , meaning that these stars are more difficult to detect .
that being said , for deeper transit surveys , such as some of the all - sky surveys , these populations may play an interesting role . in that case
, it would be fairly simple to add them into our model .
we chose the bahcall model instead of other more detailed models due to speed and portability considerations .
while others , such as the besancon models @xcite , take into account features such as density variations from the spiral structure of the galaxy and the galactic bulge , they typically require over 30 minutes to simulate a given star field . additionally , many are only accessible through webpage interfaces , and therefore difficult to include in a self - contained prediction algorithm .
we do note that comparisons between our galactic model and the besancon model revealed star count differences of only 10% in the kepler and tres fields over a range of magnitude limits .
we also model the density of interstellar dust as a double exponential @xmath117.\ ] ] which gives the dust density at a distance @xmath14 along a particular line of sight ( @xmath118 ) relative to the density in the solar neighborhood ( i.e. - @xmath119 , @xmath120 ) . to calculate the extinction in a given bandpass for a star at a given distance
, we then integrate equation ( [ eq:290 ] ) over @xmath14 to arrive at the effective path length through an equivalent solar neighborhood dust density , @xmath121 the v - band extinction in magnitudes , @xmath122 , is given by this path length multiplied by the appropriate normalization constant : @xmath123 .
we adopt @xmath124 mag kpc@xmath125 @xcite .
we determine the extinction in other bands by using the extinction ratios ( @xmath126 ) from @xcite , assuming that @xmath127 .
it is important to keep in mind that in the preceding derivations , we made several assumptions about the nature of extrasolar planetary systems , as well as some simplifications to keep the math and computing time more manageable .
_ detectability is best measured through @xmath31 : _ while s / n is a better metric than photometric precision to determine if a transit will be detectable , and the @xmath31 parameter is a good proxy for s / n , actual transit surveys use a wide variety of tests to detect transits .
many tests , such as the visual evaluation of the data by a human being , are difficult to simulate numerically .
thus , using just a @xmath31 test is a simplification , as human inspection may impose a much higher effective s / n threshold .
as we show for typical surveys , the number of detections is a strong function of the s / n threshold .
thus a proper determination of the effective s / n threshold imposed by all cuts is necessary before any robust conclusions can be drawn about the rate of detections relative to expectations .
this is a fundamental - and unavoidable - limitation of our study .
_ no stellar limb - darkening : _ we have neglected to include terms for stellar limb - darkening because , as stated above , to first - order its effects are canceled by the the effects of ingress and egress on the transit lightcurve @xcite .
_ stars are blackbodies : _ we assume that the radiated energy of a star can be described as a blackbody . while in general this is a fair approximation ( to within 0.5 mag , see figures 11 and 12 ) , it does not hold true for the bluer bandpasses we consider ( u - band in particular ) and for very low - mass stars .
_ simplified galactic structure : _ while we do treat variations in stellar density and interstellar dust densities in a general sense , our formalism is a simplified double exponential model that does not account for the spiral structure of the galaxy or patchiness in the dust distribution
. therefore detailed , accurate predictions for an individual field may be better served by incorporating more sophisticated galactic models ( e.g. @xcite ) .
_ no binary systems : _ while we only consider main sequence stars , we do not model the statistics of binary systems ; in our formulation all stars are treated as single . though this is clearly not the case , it was outside the scope of this project to go deeply into binary frequencies , the ability of planets to form in such systems , and whether or not such planets would be detectable . however
, we can make a rough estimate of the expected magnitude of the effect of binaries .
first , we assume that 1/3 of our stars have no binary companion @xcite .
we assume the remaining 2/3 of the stars are in binary systems .
a typical wide - field survey has a psf size of @xmath128 , which corresponds to a separation of @xmath129 , and a period of @xmath130 for an equal - mass solar - mass binary . adopting the period distribution for binary stars from @xcite
, we estimate that @xmath131 of the binary systems are unresolved . of these unresolved binaries ,
we conservatively assume that those with separations @xmath132 , or @xmath133 , do not host hot jupiters because the binary components are too close .
thus we estimate that @xmath134 of unresolved binaries can host hot jupiters . to estimate the detectability of transiting planets in these , consider a binary where @xmath135 , @xmath136 , @xmath137 and @xmath138 , @xmath139 , @xmath140 are the flux , radius , and mass of the primary ( `` 1 '' ) and secondary ( `` 2 '' ) , respectively .
additionally , define @xmath141 , @xmath142 , and @xmath143 .
then , for fixed planet radius and semimajor axis , assuming photon - limited uncorrelated uncertainties , the signal - to - noise ratios if the stars are resolved are @xmath144 and @xmath145 , while if the stars are unresolved the signal - to - noise ratios for transits across each component are @xmath146 and @xmath147 . assuming a constant volume density of stars and no interstellar extinction , the number of detected planets scales as @xmath148 @xcite .
thus the number of planet detections for a population of binaries in the unblended case can be written as @xmath149 , where @xmath150 is the mass ratio distribution .
similarly , for the blended case , this is @xmath151 .
using the scalings for the signal - to - noise ratios just derived , we evaluate the ratio of the number of detections in the blended case , relative to the number of detections one would expect if they were unblended , @xmath152 we adopt @xmath153 , which is roughly appropriate for the @xmath154-band ( see fig .
11 ) , and @xmath155 ( eq . [ eq:121 ] ) .
we consider two different distributions for @xmath150 .
first , we adopt @xmath150 from @xcite , which is weighted toward low mass ratio systems .
second , we consider a distribution with a substantial population of ` twins ' with @xmath156 .
specifically , we choose @xmath157 $ ] , with @xmath158 , which approximates the distribution found by @xcite .
when we then average equation ( [ eq:291 ] ) over @xmath159 , weighting by @xmath150 , we find @xmath160 for the @xcite mass ratio distribution , and @xmath161 for the @xcite mass ratio distribution .
thus , we estimate that roughly 70 - 90% of planets in unresolved binaries will still be detectable .
the basic point is that , because the mass - luminosity relation is so steep , only those binaries that have @xmath162 result in a substantial reduction in the signal - to - noise ratio .
low mass ratio companions have little affect on the ability to detect transit signals . for the mass ratio distribution of @xcite ,
these provide the bulk of the companion population , and so in this case the effects of binarity are suppressed even further .
finally , we can summarize these various components to provide a rough estimate of the fraction of planets that will be detected when we consider binary systems , relative to our fiducial assumption that all stars are single .
we find @xmath163 .
thus we expect that our estimated yields are systematically high by at most @xmath164 .
this is generally of order or smaller than other sources of uncertainty in our predicted yields .
we note that a corollary of our assertion that binaries have a small effect on the yields of transit surveys is that a substantial fraction of the planets that have been found in wide - field transit surveys should be in binary systems .
only two planets in confirmed common proper - motion binary systems have been reported @xcite , however systematic surveys for stellar companions to the host stars of the known transiting planets have not been performed , and therefore the true incidence of binary systems amongst the sample of transiting planets is not known .
using the preceding formalism , we simulated four different surveys modes used by transit searches : point - and - stare , drift - scanning , space - based , and all - sky surveys .
for the point - and - stare and drift - scanning cases , we use the tres and xo transit surveys as our examples , respectively .
both surveys have been successful in their hunt for planets , and so we are able to directly compare the results of our calculations with the actual number of detections . for vhjs and hjs , we simulate the detection of planets with @xmath165 , and with the period distribution from @xcite that we describe in section 2.2 .
c|c band & 550 - 700 nm + @xmath166 limit & @xmath167 + observation time & varies + exposure time & 90 seconds + ccd read time & 25 seconds + telescope diameter & 0.1 m + throughput & 0.6 + psf ( fwhm ) & 25 + @xmath168 & 10 + ccd full - well depth & @xmath169 + field of view & @xmath170 + galactic longitude & varies + galactic latitude & varies c|cc field & observation time & galactic coordinates + and0 & 765 hrs & [ 126.11 , -015.52 ] + cyg1 & 609 hrs & [ 084.49 , + 010.28 ] + cas0 & 628 hrs & [ 120.88 , -013.47 ] + per1 & 643 hrs & [ 156.37 , -014.04 ] + uma0 & 497 hrs & [ 168.87 , + 047.70 ] + crb0 & 537 hrs & [ 053.49 , + 048.92 ] + lyr1 & 582 hrs & [ 077.15 , + 017.86 ] + and1 & 435 hrs & [ 109.03 , -017.62 ] + and2 & 332 hrs & [ 115.52 , -016.21 ] + tau0 & 406 hrs & [ 169.83 , -015.94 ] + uma1 & 500 hrs & [ 156.32 , + 054.01 ] + her1 & 592 hrs & [ 063.26 , + 026.42 ] + lac0 & 686 hrs & [ 097.79 , -008.86 ] we demonstrate our method for space - based surveys by simulating the upcoming kepler mission , and comparing our predictions with other predictions for kepler in the literature .
similarly for the all - sky mode , we examine the future planned observing programs of lsst , sdss - ii , and pan - starrs and make predictions for each . for each mode ,
we have tried to approximate the effect of red noise on detectability by increasing our s / n cut @xcite . to show the effects of a more accurate red noise calculation , in section 4.5 we evaluate its effects on the tres survey .
point - and - stare observing is the traditional observing mode of tracking one field of sky continuously for all , or part , of the night .
there are several operational photometric transit surveys that use this particular method of observing , including hatnet @xcite , superwasp @xcite , and kelt @xcite , but we chose to model is the trans - atlantic exoplanet survey ( tres ) .
tres is composed of a network of three telescopes : stare on the canary islands , psst at lowell observatory in arizona , and sleuth on mt .
palomar , california .
all are small , 10 cm aperture , wide - field ( @xmath171 ) , ccd cameras that operate in unison to observe the same field of sky nearly continuously over one to two month periods . the survey specifics are described in more detail by @xcite and by @xcite .
the relevant quantities for our modeling are listed in table 1 .
-0.0 in we fit to the rms plots posted on the sleuth observing website to determine the background sky brightness in the tres bandpass ( 550 - 700 nm ) , using the tres team s stated photon throughput of 0.6 , and we estimate that the tres bandpass has a sky background of @xmath172 .
c|ccccccccccccccc|c & and0 & cyg1 & cas0 & per1 & uma0 & crb0 & lyr1 & and1 & and2 & tau0 & uma1 & her1 & lac0 & total + vhjs & 0.38 & 0.37 & 0.34 & 0.34 & 0.18 & 0.21 & 0.32 & 0.24 & 0.19 & 0.22 & 0.17 & 0.21 & 0.41 & 3.59 + hjs & 0.34 & 0.30 & 0.29 & 0.28 & 0.16 & 0.18 & 0.27 & 0.18 & 0.09 & 0.15 & 0.16 & 0.19 & 0.34 & 2.91 + both & 0.72 & 0.67 & 0.63 & 0.62 & 0.34 & 0.39 & 0.59 & 0.42 & 0.28 & 0.37 & 0.33 & 0.40 & 0.75 & 6.50 in our simulations , we treat the three tres telescopes as one single instrument .
this is clearly an oversimplification . while these telescopes are quite similar , there are differences between the data quality and quantity obtained from each site . for example , sleuth must contend with a higher sky background due to its proximity to urban areas , whereas the fraction of clear nights for psst is typically lower than for the other two sites
however , accounting for these differences is difficult , and furthermore their effect on our predicted detections is small compared to the other uncertainties in our model .
we therefore ignore these effects . in dealing with the window probability of the tres network ,
we were greatly aided by the plots posted on the sleuth website which showed the window probability for each field , as well as the total network observing time each field had received .
we simulated thirteen different tres fields using this methodology .
the network observation times and the location of each field are shown in table 2 .
the data for this table was collected off from the sleuth observing website .
the expected detections for each of the fields ( and the total number of detections over all simulated fields ) are shown in table 3 .
the effects of our galactic structure model can clearly be seen in distribution of expected detections .
both the uma0 and crb0 fields are substantially above the plane of the galaxy ( @xmath173 ) , and thus show a much lower detection rate than those fields directed towards denser regions of the sky .
figure 2 shows the distributions of detections over various parameters in the lyr1 field . for figure 2 , we show two distributions , one assumes white noise in the photometry , the other assumes 3 mmag of red noise .
for the white noise estimates , we used a signal - to - noise limit of s / n @xmath174 and a magnitude limit of @xmath175
. lyr1 is the field in which tres-2 was discovered , and is adjacent to lyr0 , which contains tres-1 .
we therefore have plotted the location of these two stars on the distributions shown in figure 2 .
both stars reside close to the peaks of our parent star mass and radius distributions .
the third and fourth planets discovered by tres , tres-3 and tres-4 , are not plotted in figure 2 , as they reside in the hercules fields , which are at a higher galactic latitude than lyra . also of note
is that with our assumptions half of the expected detections the f - dwarf range . among the false positives encountered by the tres team ,
the most frequent are eclipsing f / m - dwarf binaries .
figure 2 offers a partial explanation for this , showing that roughly half of tres planet - like signals come from f - dwarfs .
s / n@xmath176 & 6.06 & 5.72 & 11.78 + s / n@xmath177 & 4.68 & 4.05 & 8.73 + s / n@xmath178 & 3.59 & 2.91 & 6.50 + s / n@xmath179 & 2.77 & 2.13 & 4.90 + s / n@xmath180 & 2.16 & 1.59 & 3.75 @xmath181 & 2.13 & 1.95 & 4.08 + @xmath182 & 2.95 & 2.58 & 5.53 + @xmath183 & 3.59 & 2.91 & 6.50 + @xmath184 & 3.86 & 3.01 & 6.87 + @xmath185 & 3.94 & 3.03 & 6.97 + @xmath186 & 3.96 & 3.05 & 7.01 @xmath187 & 1.66 & 1.17 & 2.83 + @xmath188 & 2.52 & 1.90 & 4.42 + @xmath165 & 3.59 & 2.91 & 6.50 + @xmath189 & 4.53 & 3.88 & 8.41 + @xmath190 & 5.52 & 5.04 & 10.56 our predicted number of detections for the selected tres fields of @xmath191 covers only half of the fields observed by tres , so we may roughly extrapolate this result to @xmath192 detections over all of the fields . this is roughly three times the four planets that tres has actually detected , though we hesitate to draw strong conclusions from this comparison .
while we have the luxury in our predictions of imposing draconian signal - to - noise ratio and magnitude cuts in the data , in real life the tres detections are identified in data using more flexible selection effects .
a better comparison between our predictions and the actual results of tres ( or any survey ) would therefore require a better understanding of the selection effects in the observed data set .
c|c band & 400 - 700 nm + @xmath193 limit & @xmath194 + observation time & @xmath195 hrs + exposure time & 54 seconds + ccd read time & n / a + telescope diameter & 0.11 m ( two ) + throughput & 0.16 + psf ( fwhm ) & 45 + @xmath168 & 25.4 + ccd full - well depth & @xmath196 + field area & @xmath197 + galactic longitude & varies + galactic latitude & varies to see how different selection criteria change our predictions for tres , tables 4 - 6 show the effects of changing the s / n limit , the magnitude limit , and the radius of the targeted planets on the detection numbers .
a changing magnitude limit is particularly interesting , since all four of the tres detections have been around stars with @xmath198 .
this might indicate that the effective magnitude limit of the tres survey is brighter than what is described in the published literature .
if this is the case , then using a limit of @xmath198 lowers the number of expected detections in our simulated fields to @xmath199 , and total detections to @xmath200 .
-0.0 in the xo survey is a wide - angle ground - based photometric survey that uses drift - scanning for its observations .
the xo telescope is a pair of 11 cm cameras in place at the haleakala observatory in hawaii .
it monitors thousands of stars with @xmath201 in six fields spaced evenly around the sky in right ascension .
the cameras scan within these fields , sweeping in declination from @xmath202 to @xmath203 every ten minutes .
typically , this provides over 3000 observations per star in each of the xo fields .
more on the specifics of the xo system can be found in @xcite .
the relevant parameters for our predictions are listed in table 7 . to determine the throughput of xo and the background sky magnitude in the xo band ( 400 - 700 nm )
, we fit to the rms diagram shown in figure 8 of @xcite . with the addition of their stated calibration and scintillation error
, we were able to exactly reproduce the relations shown in that figure using a background sky magnitude of 19.4 mag @xmath204 and a system throughput of 0.16 .
cccccc|c 00 hrs & 04 hrs & 08 hrs & 12 hrs & 16 hrs & 20 hrs & total + 0.34 & 0.35 & 0.34 & 0.30 & 0.32 & 0.37 & 2.02 additionally , we checked to ensure that the values we used for the observation time and the number of nights observed generated an analytic window probability similar to that shown in figure 7 of @xcite .
our ability to do this was greatly aided by the xo team s description of how many nights they lost to weather and mechanical trouble .
we simulated each of the six xo fields for our predictions .
the results for each field using the base parameters of s / n@xmath178 , @xmath201 and @xmath165 are shown in table 8 .
the first planet discovered by the xo survey , xo-1b , @xcite is located in in the field at 16 hrs , and so we have also plotted the distribution of detections within this field over a variety of parameters in figure 3 .
xo-2b @xcite and xo-3b @xcite are in noticeably different fields ( 08 hrs and 04 hrs , respectively ) , so were not included on the plot .
s / n@xmath176 & 3.45 & 2.41 & 5.86 + s / n@xmath177 & 2.20 & 1.27 & 3.47 + s / n@xmath178 & 1.37 & 0.65 & 2.02 + s / n@xmath179 & 0.83 & 0.33 & 1.16 + s / n@xmath180 & 0.50 & 0.17 & 0.67 @xmath205 & 1.09 & 0.55 & 1.64 + @xmath206 & 1.37 & 0.65 & 2.02 + @xmath207 & 1.50 & 0.69 & 2.19 + @xmath208 & 1.54 & 0.70 & 2.24 + @xmath209 & 1.55 & 0.71 & 2.46 @xmath187 & 0.29 & 0.09 & 0.38 + @xmath188 & 0.69 & 0.25 & 0.94 + @xmath165 & 1.37 & 0.65 & 2.02 + @xmath189 & 2.07 & 1.18 & 3.25 + @xmath190 & 2.93 & 1.92 & 4.85 tables 9 - 11 show the effects of changing the s / n limit , the magnitude limit , and the radii of the targeted planets on xo s predicted detection numbers . for our base case of @xmath201 and s / n@xmath178 , we expect that xo will detect 2.02 planets .
if we consider the three actual detections from xo , it seems that , similar to tres , the actual s / n and magnitude limit of the xo survey may differ from what is in the literature .
for example , all three xo parent stars are at @xmath205 ( xo-1 , the dimmest , is at @xmath210 ) . to see how this affected our calculated distributions , we also simulated the field at 16 hrs assuming s / n@xmath176 , @xmath201 , and s / n@xmath176 , @xmath205 .
the results are included on figure 3 . in these cases ,
the total number of detections increases to 5.86 and 4.21 , respectively .
to make predictions for a space - based mission , we chose nasa s kepler mission , which is primarily designed to look for earth - sized extrasolar planets .
it consists of a single spacecraft that is currently scheduled to be launched into an earth - trailing heliocentric orbit in early 2009 .
onboard , kepler will have 42 ccds that will measure the light from 100,000 main - sequence stars within the telescope s field of view .
ideally , kepler will identify planets orbiting within the `` habitable zone '' of their parent stars which would be capable of possessing liquid water on their surfaces .
c|c band & 400 - 850 nm + @xmath211 & 0.86 + @xmath193 limit & @xmath212 + observation time & 4 years + exposure time & 3 seconds + ccd read time & negligible + telescope diameter & 0.95 m + throughput & .3 + psf ( fwhm ) & 10 + @xmath168 & 3.98 + ccd full - well depth & @xmath213 + field of view & @xmath214 + galactic longitude ( new ) & @xmath215 + galactic longitude ( old ) & @xmath216 + galactic latitude ( new ) & @xmath217 + galactic latitude ( old ) & @xmath218 beyond looking for extrasolar earths , the immense amount of photometry that kepler is expected to produce over its four year operational lifetime will also allow for the detection of larger jovian worlds transiting other stars .
previous estimates of the number of vhjs and hjs expected to be found by kepler have predicted the discovery of about 180 hjs ( no vhjs had been detected at the time , and so they were not treated ) @xcite .
clearly , given that there are only 46 currently known transiting planets , this would be a significant increase in the number of known planets to simulate kepler with our methodology , we used the values in table 12 to simulate kepler observations out to a distance of @xmath219 , using a signal - to - noise ratio of sn@xmath220 .
we gathered the information in table 12 from both the kepler mission website as well as published descriptions of the satellite @xcite .
three items are of note .
first , our value for the photon detection efficiency of kepler is based upon information from the kepler website that a 12th magnitude g2 dwarf will show @xmath221 .
secondly , the value we use for the extinction ratio in the kepler bandpass , @xmath222 , is an average that we calculated by integrating the extinction curve over the kepler bandpass ( 400 - 850 nm ) @xcite .
finally , the kepler website and much of the literature states that kepler will use a dim magnitude cutoff of @xmath223 . in reality
, there is no `` hard '' magnitude cutoff at which stars will be excluded from observations .
indeed , kepler is expected to observe promising stars that are much dimmer than @xmath224 ( d. w. latham , private communication ) .
therefore , for our more explicit calculations of the expected number of hj and vhj detections for kepler , we adjusted the magnitude limit such that the number of stars in the field agreed with the expected number of kepler targets ( 100,000 in the case of the new field ) . for other , more general , simulations
, we used several different magnitude limits to determine what effects these had on kepler s expected detections .
while the location of the kepler field is currently centered on galactic coordinates @xmath225 , up until 2004 it was instead situated closer to the galactic plane at @xmath226 .
the switch in the position of the field to higher galactic latitudes was done in order to avoid sorting through the larger number of giant stars that are found in fields directed along the plane of the galaxy .
much of the previously published literature examining the kepler mission have used the coordinates of the old field . for the purpose of comparison
, we therefore simulated both the old and new kepler fields .
-0.15 in -0.15 in looking in the new kepler field with a magnitude limit of @xmath223 , we simulated the number of detections as a function of orbital period for jupiter- , earth- , and mars - sized worlds , assuming that every star possessed a planet with the given period and radius ( figure 4 ) , and that red noise was negligible .
interestingly , the difference between the number jupiter- and earth - sized planets detected is minimal for @xmath227 days , which indicates that kepler is able to detect all the planets within this radius range , assuming that they transit .
therefore , for @xmath228 and @xmath227 days , the s / n of transiting jupiter- and earth - sized planets is large enough that kepler s s / n limit is irrelevant , and the @xmath223 magnitude cutoff is the limiting factor in the number of detections . for smaller planets the size of mars ,
the s / n of their transits are in the neighborhood of the kepler s / n limit for the brightest stars and shorter periods , and so mars - sized worlds are detected at a much lower rate .
-0.0 in -0.1 in to more accurately characterize the magnitude- and s / n - limited detection regimes , we plotted the number of detections as a function of planetary radius for orbital periods of 1 day and 365 days ( figure 5 ) .
in addition in a magnitude limit of @xmath224 , we also plot the same relations for a limit of @xmath229 .
one can clearly see the flattening of these curves as the radius of the target planets increase , indicating the increasing effect of the magnitude limit on the detection numbers .
as is expected in the s / n limited regime , detections go as @xmath230 @xcite , though at the lower end saturation of the kepler ccds serves to cut out some of these detections .
one of the major goals of the kepler mission is the detection of planets in a star s habitable zone that would be capable of supporting life .
the habitable zone is defined as the distance from the parent star at which liquid water water could exist on a planet s surface .
we estimate the detection rate of habitable zone planets by assuming that all stars have a planet with a semi - major axis equal to @xmath231 this assumption is simple and likely optimistic . in reality
, the habitable zone has a finite width , and habitable planets will have a range of radii .
thus , a better estimate would be to adopt a distribution of planet masses and radii , and integrate over the range of radii and periods considered habitable .
furthermore , the assumption of one habitable planet per star is probably optimistic , and so the actual number of detected habitable earths will be smaller than we estimate , depending on the real frequency of earth - like planets . for the sake of simplicity , however , we will adopt the assumptions above . with this assumption , we simulated the detection of habitable earths in both the old and new kepler fields ( figure 6 ) .
surprisingly , despite there being nearly twice as many main sequence stars in the old field , the number of detected habitable earths remained nearly the same in the two fields .
this occurred because at the magnitude relevant for kepler , the increased star count in the old field was mainly due to the larger number of a and f main - sequence stars found closer to the galactic plane ; habitable planets are doubly hard to detect around these stars due to the lessened depth of transit , and the longer periods required for a planet to be in the habitable zone .
the exact distribution of stars by spectral type for the old and new fields is given in table 13 , and is discussed further below . for the 106,000 stars in the new kepler field , we find that kepler will be able to detect 79 habitable earth radius planets , assuming that every star posses one .
-0.125 in -0.15 in in general , figure 6 provides an excellent example of the benefit of repositioning the kepler field away from the galactic plane .
namely , the number of large dwarfs and giant stars ( which we have not treated ) found in the galactic plane can be avoided , reducing the false - positive rate , while the number of habitable earths detected remains nearly the same @xcite .
it is also possible that life may arise on non - earth - sized worlds .
a world the size of mars might develop quite differently than it has in our own solar system if it were closer to its parent star , and jovian worlds in the habitable zone could possess large moons capable of supporting life .
figure 7 shows the number of detections of jupiter- , earth- , and mars - sized habitable worlds , assuming that every star has a planet , against a variety of parameters in the new kepler field . for completeness
, the magnitude limit was set to @xmath232 .
one can see in the mass and radius plots an explicit demonstration of why changing the number of bright , massive main - sequence stars in the field does not effect the number of habitable planet detections . as was mentioned earlier all detections
drop to zero beyond @xmath233 . in the plot showing the number of detections binned against the absolute v - magnitude of the parent star we have additionally shown the effect of changing the apparent v - magnitude limit on the distribution of earth - sized planets . using the hj and vhj frequencies provided by @xcite
, we were able to make explicit predictions about the number of these worlds that kepler should find . as mentioned previously , @xcite ( hereafter jd03 ) have made their own estimates of kepler s hj detection ability ( as noted previously , no vhjs were known at the time , and so were not considered ) , thus providing a useful point of comparison . as before , we used a limit of s / n@xmath220 , though for jupiter - sized planets kepler is magnitude , and not signal - to - noise , limited .
-0.0 in -0.0 in ccccccc|c & b & a & f & g & k & m & all + jenkins & doyle : & 11014 & 36708 & 81085 & 61347 & 15573 & 831 & 206558 + calculated ( old field ) : & 4278 & 31370 & 76630 & 62891 & 18243 & 1264 & 194678 + calculated ( new field ) : & 338 & 5066 & 41094 & 45568 & 13351 & 942 & 106342 ccccccc|c & b & a & f & g & k & m & all ( hjs ) + jenkins & doyle : & 10 & 32 & 71 & 54 & 14 & 1 & 181 ( 181 ) + calculated ( old field ) : & 4.2 & 27.3 & 55.5 & 33 & 8 & 0.4 & 127 ( 72.8 ) + calculated ( new field ) : & 0.3 & 4.4 & 28.6 & 23.5 & 5.8 & 0.33 & 62.9 ( 36 ) at the time of the jd03 estimate the kepler field was centered on its old coordinates . for the purposes of comparison , we therefore examined the old kepler field in addition to the new field .
tables 13 and 14 show the distribution of our predictions by spectral type . to facilitate comparison with jd03 ,
we have included not only the final prediction numbers for both old and new kepler fields , but also the distribution of the calculated total number of stars in each field .
as stated above , in both fields we set the magnitude limit of the simulation such that the total number of stars in the fields agreed with the number of expected kepler targets . in the case of the new field , this is 100,000 stars , while for the old field we matched the number of stars predicted by jd03 .
this corresponded to a magnitude limit of @xmath234 in the new field and @xmath235 in the old field .
looking at the detections , one can see in table 14 that our final prediction of @xmath236 for the old kepler field is lower than the jd03 prediction .
this is primarily a result of the fact that the @xcite results that we use for @xmath85 ( see section 2.2 ) posit a hj and vhj frequency approximately @xmath237 of the frequency used in jd03 .
an important difference , however , between our work and jd03 is our consideration of vhjs , which were not treated by jd03 .
this has the effect of `` doubling '' the number of planets we expect kepler to detect ( as demonstrated by the final values in parentheses in table 14 ) .
we therefore arrive at detection rates for vhjs and hjs that are 2/3 that of jd03 for the old kepler field , and 1/3 that of jd03 in the new kepler field . for the old field , we predict 54.2 vhjs and 72.8 hjs . for the new field ,
we predict 26.9 vhjs , and 36.0 hjs . in the new kepler field , we see that looking away from the plane of the galaxy reduces the number of hjs and vhjs detected by half to @xmath238 , which matches the overall drop in the star count in the new field .
the distribution of detections changes substantially , shifting more towards f- , g- , and k - dwarfs .
this is the caused by our model of the galaxy s vertical structure , which uses reduced scale heights for earlier stars .
thus , most of the difference in star counts between the old and new fields is caused by fewer b- , a- , and f - dwarfs , leading to proportionally smaller detection numbers around these types of stars .
while the prediction number of detections in the new kepler field is lower than the corresponding jd03 prediction would be for the same star count , this is accounted for in the same way as in the old field .
indeed , we find that our predictions are consistent with those of jd03 , except for the difference in our assumption about the frequency of hjs and vhjs .
the eventual results from kepler therefore offer an excellent opportunity to more robustly determine the actual prevalence of short period jovian worlds .
c|cccc survey & lsst & sdss - ii & pan - starrs & pan - starrs wide + @xmath193 limit ( sun - like ) & 18.51 & 15.56 & 14.99 & 12.49 + @xmath193 limit ( m - dwarfs ) & 23.13 & 20.18 & 19.61 & 17.11 + observation time & 10 years & 37.5 days & 5 months & 5 months + telescope diameter & 6.5 m & 2.5 m & 1.8 m & 1.8 m + throughput & 0.5 & 0.5 & 0.5 & 0.5 + field of view & @xmath239 & @xmath240 & @xmath241 & @xmath241 + area surveyed & @xmath242 & @xmath243 & @xmath244 & @xmath245 in addition to wide - field photometric surveys , several current and planned projects aim to repeatedly gather photometry over gigantic swathes of the sky . while not all of them have extrasolar planet detection as one of their goals , the large areas covered by these surveys , together with their long durations ( from 3 to 10 years ) , offer the possibility of detecting hundreds of transiting vhjs and hjs .
we examined three all - sky surveys , lsst , sdss - ii , and pan - starrs .
the large synoptic survey telescope ( lsst ) will be a large 8.4 m telescope , with an effective aperture of 6.4 m , situated in northern chile .
it is scheduled to finish construction sometime in 2014 . using a 9.6 deg@xmath246 field of view
, the lsst team intends to image 20,000 deg@xmath246 of the sky repeatedly over the course of 10 years .
sdss - ii is the second phase of the sloan digital sky survey ( sdss ) which began in june , 2005 and is scheduled to last for three years until june of 2008 .
it is composed of three distinct surveys , the sloan legacy survey , segue , and the sloan supernova survey @xcite .
the supernova survey is the only one of these three that will collect photometry capable of finding transiting planets ; the other two survey modes are focused on non - variable objects within and outside the galaxy . to execute the supernova survey
, sdss uses a 2.5 m telescope on apache point in new mexico with a 6.25 deg@xmath246 field of view . over the three year run of sdss - ii
, the team expects the supernova survey to image about 300 deg@xmath246 of the sky near the galactic poles every two nights for the three months a year the poles are visible .
given the observing cadence the survey expects to use , this translates into a total observing time of 37.5 days .
@xmath247 & 4.052 & 0.027 & 0.800 + lsst ( 18.51 ) & 1.609 & 0.027 & 0.387 + @xmath248 & 1.125 & 0.026 & 0.293 + @xmath249 & 0.219 & 0.025 & 0.087 + sdss - ii ( 15.56 ) & 0.146 & 0.020 & 0.063 + pan - starrs ( 14.99 ) & 0.083 & 0.016 & 0.041 + @xmath228 & 0.029 & 0.009 & 0.017 + pan - starrs wide ( 12.49 ) & 0.005 & 0.003 & 0.004 + @xmath250 & 0.003 & 0.002 & 0.002 lsst ( 23.13 ) & 2.7558 & 0.0888 &
0.7764 + sdss - ii ( 20.18 ) & 0.2473 & 0.0398 & 0.1136 + @xmath247 & 0.2081 & 0.0368 & 0.0989 + pan - starrs ( 19.61 ) & 0.1422 & 0.0304 & 0.0725 + @xmath248 & 0.0257 & 0.0105 & 0.0169 + pan - starrs wide ( 17.11 ) & 0.0092 & 0.0048 & 0.0068 + @xmath249 & 0.0047 & 0.0015 & 0.0028 + @xmath228 & 0.00017 & 0.00015 & 0.00016 + @xmath250 & 0.00001 & 0.00001 & 0.00001 finally , pan - starrs - the panoramic survey telescope and rapid response system - will be a group of four 1.8 m telescopes atop the summit of haleakala , hawaii @xcite .
the first telescope , ps-1 , is nearing completion .
when fully operational , pan - starrs will embark on a 3 year - long `` medium deep '' survey that will observe 1200 square degrees of the sky near the galactic poles for a total of five months .
we also looked at a notional `` wide '' pan - starrs survey that looked at 10x the amount of sky as the medium deep survey , but used the same amount of observing time .
to characterise each of these surveys and find how many transiting vhjs and hjs they should discover , we first calculated the v - magnitude limits at which each of the surveys will be able to detect transiting jovian worlds .
the complete derivation of this limit is given in appendix e. table 15 shows the results of these calculations , as well as the other relevant parameters for each of the surveys .
we calculated magnitude limits for detecting giant planets around both sun - like ( @xmath251 ) stars and m - dwarfs ( @xmath252 ) for s / n@xmath176 .
-0.1 in -0.1 in to translate these magnitude limits into the number of transiting planets each of the surveys should detect , we used our methodology to calculate the density of vhjs and hjs for various limiting magnitudes , in numbers per square degree , along the galactic plane , at the galactic poles , and averaged around the sky .
table 16 shows the resultant values for a series of fiducial magnitude limits , as well as the planet densities at the magnitude limits of our modeled surveys for sun - like stars ; table 17 shows transiting planet densities around m - dwarfs .
additionally , in figure 8 we show the transiting planet density around sun - like stars and m - dwarfs as a function of galactic latitude .
it is interesting to note the dip near the galactic plane , which occurs as a result of extinction from interstellar dust . knowing the angular size of each survey
, we use these densities to predict the number of planets that will be seen . for sdss - ii and pan - starrs , which are both pointed towards the galactic poles
, we used the polar densities , while for lsst and the wide pan - starrs , with their much larger and wide - ranging survey area , we used the all - sky average density .
the final predictions for both sun - like stars and m - dwarfs are shown in table 18 .
we note that these predictions are based on a simpler model than what we have used in other portions of this paper .
for instance , we do not include the effects of ccd saturation or sky - background .
nevertheless , our calculations should provide a good expectation for the range of the detection numbers that these all - sky surveys can expect .
the large difference between the number of planets expected to be detected by lsst and the other surveys is caused by two effects .
first , sdss - ii and pan - starrs cover much smaller areas of the sky to shallower depths than lsst , and second , lsst is not primarily directed towards the galactic poles , greatly increasing the density of stars for which planet detection is possible .
pan - starrs wide , meanwhile , has a substantially brighter magnitude limit than any of the other surveys , which prevents it from detecting as many planets as lsst , despite the similarities in the size of their survey areas .
pan - starrs wide does detect more planets than the planned narrow survey , primarily because the latter `` runs up '' against the scale height of the galaxy .
c|cc & sun - like ( g2 ) & m - dwarfs ( m5 ) + lsst & 7740 & 15530 + sdss - ii & 6.0 & 11.9 + pan - starrs & 19.2 & 36.5 + pan - starrs wide & 48.0 & 81.6 the real difficulty , however , in managing the 7,740 or 15,530 detections that we expect lsst to make will be properly identifying and separating the false - positives from the actual planetary transit events .
indeed , to date the wide - angle surveys have generated approximately 15 false positive candidates for every actual transiting planet they have seen ( d. w. latham , private communication ) .
if we extend this ratio to lsst , sdss - ii , pan - starrs and pan - starrs wide , we would expect them to create 116,000 , 90 , 290 , and 720 false positives , respectively around sun - like stars , and 233,000 , 180 , 550 , and 1,200 around m - dwarfs , respectively .
working through this many transit candidates ( especially in the case of lsst ) is an intensive , time - consuming task .
we would therefore expect the all - sky surveys , despite the high absolute number of predicted detections , to discover transiting planets at about the same rate as the operational wide - angle surveys , as the identification rate is predominantly limited by the ability to make follow - up observations of candidates to determine whether or not it is a false positive . as pointed out by @xcite , the general mathematical description we have used for the s / n of a transit , equation ( [ eq:30 ] ) , assumes uncorrelated errors on the collected photometric data points ( white noise ) . in practice this may not be the case , as phenomenon such as changes in the seeing , airmass , telescope tracking , etc . introduce systematic trends into the observed data that are correlated on the timescales of transits ( red noise ) .
this has the effect of substantially decreasing the effective s / n displayed by a transit , thus making detections more difficult .
specifically , @xcite show that to account for red noise , equation ( [ eq:30 ] ) must be modified by adding a term to include a covariance coefficient @xmath253 between the @xmath254 and @xmath255 measurements , @xmath256 this is equation ( 4 ) in @xcite , using the variables we previously defined for our equation ( [ eq:30 ] ) .
the authors give a more manageable version on page 8 of their work , which relates the s / n with red noise to the s / n without red noise : @xmath257,\ ] ] where @xmath258 is the white noise @xmath31 statistic we calculate in equation ( [ eq:30 ] ) , @xmath259 is the number of data points in the single @xmath260 transit ( which on average is just @xmath261number of transits ) , and @xmath262 is the red component of the noise in the data .
c|ccccccccccccc|c & and0 & cyg1 & cas0 & per1 & uma0 & crb0 & lyr1 & and1 & and2 & tau0 & uma1 & her1 & lac0 & total + vhjs & 0.34 & 0.33 & 0.31 & 0.30 & 0.17 & 0.20 & 0.30 & 0.22 & 0.17 & 0.20 & 0.16 & 0.20 & 0.36 & 3.26 + hjs & 0.25 & 0.22 & 0.21 & 0.21 & 0.11 & 0.13 & 0.19 & 0.12 & 0.07 & 0.11 & 0.11 & 0.14 & 0.25 & 2.15 + both & 0.59 & 0.55 & 0.52 & 0.51 & 0.28 & 0.33 & 0.49 & 0.34 & 0.24 & 0.31 & 0.27 & 0.34 & 0.61 & 5.41 [ cols="^ " , ] to properly use this formulation , we must therefore have some idea of what value @xmath262 takes .
@xcite infer values of 3 mmag of red noise for typical ground - based surveys , which can be reduced to approximately 2 mmag after applying de - trending algorithms to the data .
we therefore took this tentative value of @xmath263 and used it to re - simulate the tres fields to see what effect red noise would have on detection numbers .
as noted in @xcite , we set a signal - to - noise ratio limit of @xmath264 , which is the @xmath265 value of the detection of tres-1b ; tres-2b has a higher value of @xmath266 .
the results are shown in table 19 .
figure 2 shows ( in addition to tres with only white noise ) the distribution of detections using 2 mmag of red noise and @xmath264 .
c|c @xmath262 & hab .
earths + @xmath267 & 14.37 + @xmath268 & 32.53 + @xmath269 & 72.88 + 0 & 78.89 compared to a white s / n limit of @xmath270 , using @xmath271 mmag and @xmath264 drops the total number of expected detections by about 20% . looking at the split between hjs and vhjs ,
the hj population is disproportionately effected , which is a result of the relatively fewer transits hjs show , lowering the overall s / n of their transit signal .
similarly , figure 2 shows that red noise also cuts out the detections around higher mass stars , as these have a lower s / n overall . without de - trending the data ,
a red noise of 3 mmag would cause an even further reduction in detections , down 60% as compared to white noise .
we also examined the possible effect that red noise would have on the yield of habitable earth - radius planets from kepler mission . given that kepler is a space - based mission designed to detect transit signals with fractional depths at the level of @xmath272 , we would expect that the amount of red noise due to instrumental and environmental sources would be small , if it exists at all . however , this is generally difficult to predict before launch
. furthermore , there may exist significant astrophysical red noise due to intrinsic variability .
we therefore estimated the expected kepler yield for a set of three values for @xmath262 , assuming that every star has an earth - radius planet with @xmath273 ( table 20 ) .
for a relatively large amount of red noise at @xmath274 , we find the number of habitable earths - radius is drastically reduced from our baseline ( uncorrelated noise ) prediction , from 78.89 down to 14.37 .
this is not surprising , since earth passing in front of a sun - like star has a transit depth of only @xmath275 . for this amount of red noise ,
kepler is restricted to detecting earth - like planets only around k- and m - dwarfs .
this shifts the expected distribution of host stars towards dimmer magnitudes ( figure 6 ) . for smaller values of @xmath276 and @xmath277 ,
the number of habitable earths detected is reduced to 32.53 and 72.88 , respectively .
both of these estimates also assume that every target star has an earth - sized planet in the habitable zone , and therefore must be viewed as upper limits .
thus we conclude that kepler s yield of habitable zones of habitable planets will be not be strongly affected by red noise , provided that the red noise is @xmath278 on the timescales of the transits , which is @xmath279 for solar - type stars .
if the red noise is found to be larger than this , it may compromise the primary kepler science by significantly reducing the number of expected habitable planet detections .
for example , if the red noise is @xmath280 , and the frequency of habitable planets with @xmath281 is @xmath282 , there is a @xmath283 chance that kepler will not detect any .
the effects of red noise on the habitable planet yield can be mitigated by including a larger number of k and m dwarfs in the target sample .
of course , it will also be possible to extrapolate the frequency of shorter - period or more massive planets , where the effects of red noise are mitigated , into the habitable zone .
in this paper we have developed a method to more accurately predict the number of transiting planets detectable by wide - field ground- and space - based photometric surveys .
we have taken into account such factors as the frequency of gas giants around main sequence stars , the probability of transits , stellar density changes from galactic structure , and the effects of interstellar extinction .
we adopt a s / n detection threshold criterion to approximate the detectability of planetary transits . to test our model , we compared our predictions to two operating photometric surveys , tres and xo . for both , we calculated the number of statistically expected detections , as well as the population characteristics of the host stars . for tres ,
we overestimate the number of detections ( thirteen versus the four actual detections ) , while for xo our simulations yield approximately the actual number of detections ( two predicted versus three actual ) .
the exact reason for the discrepancy between our predictions and the actual tres results could be one of many . as we discuss in the tres section , from the properties of the known tres planets , it may be that the survey has a brighter magnitude cut off than the one we use in our fiducial case . if this were the case , and we changed the magnitude limit from @xmath284 to @xmath198 , the number of expected detections would drop to 8.16 from 13 .
red noise also has the potential to substantially affect tres ; a specific estimate of the magnitude of red - noise in the tres survey would be worthwhile . indeed , it is generally hard to draw specific statistical conclusions from the comparison of our results and those of the transit surveys because they typically do not adopt the strict detection criteria we have used .
promising planet candidates are often followed up even if they are beyond the stated s / n or magnitude limits of the survey .
understanding and quantifying how the survey teams select candidates is vital to appropriately deriving the statistical properties of extrasolar planets .
indeed , tables 4 - 6 and 9 - 11 demonstrate that the actual predictions depend crucially on the specific magnitude and s / n threshold used .
as more transiting planets are discovered , these statistical properties are increasingly becoming the frontier of research , shifting the focus of the field away from individual detections .
given that transiting planets are so far one of the only ways that we may infer the radius of extrasolar planets and their exact orbital properties , the statistical characteristics of this group of objects is one of the only foreseeable ways that the areas of planetary interiors , system dynamics , migration , and formation will acquire more data .
in the future , besides using the preceding formalism to design more efficient transit surveys , we would hope that a more systematic approach towards transit surveys will allow this model to be used to make more specific statistical comparisons . as the number of known transiting planets grows
, we would also expect that this formalism will be used to test different distributions of planet frequencies , periods , and radii against those observational results .
this will allow us to better understand the statistics of extrasolar planetary systems , improve our ability to find new planets , and help to understand the implications of the ones that have already been detected .
we would like to thank the referee for a helpful report .
we would also like to thank dave charbonneau and francis odonovan for providing useful discussions and information about the tres survey , peter mccullough for his correspondence about the xo survey , and scott schnee for discussions about interstellar dust .
tgb would like to thank dave latham and josh winn for helpful discussions and encouragement .
the photon luminosity in a bandpass of a given star of a given mass is calculated exactly by dividing the planck radiation law for wavelength by the energy of a photon at that wavelength , multiplying the result by the surface area of the star , and then integrating over the wavelength range of the relevant bandpass : @xmath285 where @xmath286 is the planck law , and @xmath287 is the energy of a photon .
note that this assumes that stellar radiation can be approximated as a blackbody , a crude assumption that still suffices for our purposes , as we show below .
because the widths of the bandpasses in the visible and near - infrared that we are considering are generally narrow with respect to @xmath288 , we can safely approximate this integral by the product of the central wavelength of each bandpass @xmath289 and a bandpass width @xmath290 : @xmath291 - 1},\ ] ] which , once we select a bandpass , only leaves the temperature of the source star @xmath292 to be determined . using the stefan - boltzmann law , we may relate the temperature of the source star to its bolometric luminosity and radius : @xmath293 where @xmath294 and @xmath295 are the radius and the bolometric luminosity of the source star , respectively .
we chose to treat both as functions of the mass of the source star , as this is an independent variable that allows the use of the present day mass function to describe the general stellar population . using stellar data from @xcite and @xcite , we fit broken power laws for both the @xmath296 and @xmath297 relations . for the mass - radius relation ,
we fit for two breaks in the power law , while the fitting for the mass - luminosity relation was accomplished using only one break .
the relations that we thus found are : @xmath298 @xmath299 equations ( [ eq:121 ] ) and ( [ eq:131 ] ) are plotted against the @xcite and @xcite data in figure 9 . the real test
, however , is how well these relations allow for the calculation of the effective temperature of a star at a given mass using ( [ eq:111 ] ) , as this feeds directly into the determination of the photon luminosity in a given band .
figure 10 shows a plot of our mass vs. temperature relation along with data points from @xcite and @xcite .
additionally , we also show the analytical relation provided by @xcite .
one can see that our formulation of ( [ eq:121 ] ) and ( [ eq:131 ] ) affords a close approximation to both the data and the relation provided by @xcite . substituting back into ( [ eq:101 ] ) allows us to calculate the photon luminosity of the source star , as well as the absolute magnitude of the source star in various bandpasses .
if we compare these calculated absolute magnitudes in different bands for various stellar masses against empirical magnitudes for those same masses ( figures 11 and 12 ) we can see that the two relations are similar , to within half a magnitude , especially in longer wavelengths .
shorter wavelength bands such as u , b , and to a certain extent v , have calculated magnitudes that diverge from the empirical relation below about a solar mass , which is an unfortunate effect of using blackbody relations to determine the photon luminosity , as in these higher energy bands stellar spectra increasingly do not look like black bodies .
nevertheless , the blackbody assumption is sufficiently accurate for our purposes .
roughly , if we ignore the effects of aliasing , we may say that the window probability of seeing a minimum of @xmath300 transits of a planet with orbital period @xmath16 over the course of @xmath301 nights may be broken up into four different regimes . first , if the average number of transits during the total time during which the target star is observed exceeds @xmath300 , then it is statistically `` certain '' that the transit will occur during the times of observation .
second , it may be that the entire length of the survey is long enough that there are at least @xmath300 transits , but that during the total time spent observing the average number of transits is less than that required for detection .
then one is not certain to detect @xmath300 transits , and there is a decreasing window probability ( for increasing planet period ) .
third , one may have @xmath300 transits during an observing run , but only if the first transit occurs sufficiently early during that particular run .
the probability of detecting the system then decreases at an even greater rate as the orbital period increases .
finally , the orbital period of system may be so great that there is no possible way in which one could see the minimum number of transits required for a detection , and @xmath302 .
mathematically then , @xmath303 where @xmath304 day .
@xmath305 is the average number of transits that will be seen during the time spent observing , and is given simply by : @xmath306 @xmath307 is defined as the total number of `` good '' observing nights in a particular run , and may mathematically be described as @xmath308 , where @xmath309 is the number of nights lost due to weather , technical problems , or other issues .
@xmath310 is the length of each night s observing time .
alternatively , one may simply view the quantity @xmath311 as the total amount of good observing time in a given run . to test the accuracy of our formulation , we integrated and found the average window probability for both the analytic and exact formulations linearly averaged over several period ranges ( 1 - 3 , 3 - 5 , and 5 - 10 days ) .
these are shown as a function of total observing nights under different viewing scenarios in figure 13 .
one can see that as a result of the analytic solution not taking into account aliasing , there is divergence between the two solutions as the analytic form approaches unity , as well as in the cases requiring a minimum of three transits .
a review of the literature shows that while there is agreement that the pdmf beyond @xmath312 can be treated as a power law with a slope of @xmath313 , below a solar mass this agreement breaks down .
this is because the low - mass end of the mass - function is unfortunately not as well constrained as it is for higher mass stars , reflecting the difficulty of observing dim m - dwarfs , as well as uncertainties in the mass - luminosity function @xcite .
for instance , @xcite , who used the hipparcos data sets along with their own data from the palomar / michigan state university ( pmsu ) survey to create a volume - limted sample out to 25 pc , propose a traditional power law formulation of the pdmf : @xmath314 with @xmath315 for @xmath316 , and @xmath317 for @xmath318 , and @xmath319 as the correct normalization . in a recent review ,
meanwhile , @xcite argues for a log - normal function below a solar mass : @xmath320,\ ] ] with @xmath321 , @xmath322 , and @xmath323 . for comparison ,
if we normalize both to the @xcite normalization for 5 gyr of @xmath324 and plot the two functions against each other ( figure 14 ) , we can clearly see the divergence both numerically and functionally of these two forms below a solar mass .
interestingly however , for most surveys this divergence makes little difference in the final number of predicted detections : typically reid s relation gives only 3% fewer detections than chabrier s does , if both are normalized at a solar mass .
that this difference is so small comes from the fact that most wide - angle surveys are not sensitive to detections around stars with @xmath325 .
hence , the region of greatest uncertainty in the pdmf is left unseen by these surveys . since the difference between the two relations is small , and because the data set used by @xcite is much more recent than that used by @xcite , we therefore chose to use the @xcite form of the pdmf . for the purpose of comparing the reid and chabrier pdmfs in figure 14 , we normalized both to the value given by @xcite .
going forward , however , we change the normalization to the integrated mass density found observationally by @xcite of @xmath108 .
the values for the scale length and scale height of the galactic disk that we use in our modeling are @xmath115 and @xmath114 as a function of absolute magnitude .
specifically , we calculate the vertical scale height of the disk as : @xmath326 which is a modified form of the relation found in @xcite , who instead used a value of 320 pc for the scale height of stars with @xmath327 . the large uncertainties in the value of the scale length of the disk reflects the uncertainty in the relevant literature . in their initial paper ,
@xcite use @xmath328 . since then various sources have proposed an even @xmath329 @xcite ,
have covered all the bases by using several values between @xmath330 @xcite , and have given the very exact result of @xmath331 @xcite . to see what effect different scale lengths have on our final detection numbers , we simulated observations of a @xmath332 field along the galactic plane at intervals of @xmath333 in galactic longitude , as well as fixing the galactic longitude at certain points ( @xmath202 , @xmath334 , and @xmath335 ) and varying the latitude of the fields .
figure 15 shows these results for scale lengths of 2 , 2.5 , and 3 kpc . at the extremes ( directly towards or directly away from the galactic core ) the 2 and 3 kpc scale lengths differ by only 0.15 detections per field .
we chose 2.5 kpc not only to split this difference , but also because more recent results place the scale length much lower than earlier 3 kpc estimates [ e.g. - @xcite ] . in the case of the scale height of the galactic disk
, the literature divides between those who derive the scale height from optical star counts and those who rely on near - infrared observations . in the optical case ,
the scale height is presented as a function of absolute visual magnitude that varies from 90 pc to values between 250 - 320 pc .
the near - infrared observations , on the other hand , lead to one single value of the scale height , generally around 290 pc .
this happens because measurements in the infrared are dominated by older m - dwarfs that have dispersed in an even fashion away from the galactic plane . to compare these different estimates ,
we have plotted the number of detections in a @xmath332 field along @xmath336 as a function of galactic latitude in figure 16 .
we used two of the optically - derived scale height relations , as well as a constant scale height of 290 pc .
one of the optical relations , the bahcall model , is given by @xmath337 the other optical model is taken from allen s astrophysical quantities ( aaq ) @xcite and is : @xmath338 also plotted in figure 16 are the number of detections above and below @xmath339 for the three different scale height relations , as well as the residuals of the constant 290 pc and the aaq relation against the bahcall formulation .
the exact choice of how the galactic scale height is treated leads to rather different final predictions .
specifically , while the optical relations are similar , the constant 290 pc scale height from near - infrared observations leads to substantially more detections around stars with @xmath340 . for our predictions
, we have adopted the form of the @xcite and allen s astrophysical quantities scale height relations .
we find the gradated scale height in these models appealing , because it accounts for the fact that more luminous stars with shorter lifetimes should not be dispersed far away from star - forming regions in the plane of the galaxy .
the constant 290 pc relation , because it is based on near - infrared data , reflects the sensitivity of those observations to less massive stars that are brighter in longer wavelengths .
nevertheless , we have used the 290 pc result as our value for the scale height of late - type stars with @xmath341 , in contrast to the values used in the aaq and bahcall models .
as per equation ( [ eq:30 ] ) , we have @xmath342 for the all - sky surveys that we are considering , the number of in - transit data points observed is simply the number of data points acquired in each field multiplied by the fraction of time a planet will spend in transit during its orbit : @xmath343 where @xmath344 is the total observation time of the survey , @xmath345 is solid angle of the survey telescope s field of view , and @xmath346 is the total solid angle of the sky covered by the survey . for poisson , source - noise dominated errors , we have @xmath348 where @xmath349 is the total number of photons from the source collected during @xmath47 .
we can then write , @xmath350 where @xmath351 is the fiducial fractional photometric uncertainty achieved for a star with apparent magnitude @xmath352 using a telescope with diameter @xmath353 and an exposure time of @xmath354 .
we have defined @xmath355 as the efficiency of the observing setup relative to the fiducial case . if we rearrange equation ( [ eq:480 ] ) to @xmath356 }
+ m_{v,0}\ ] ] we may now substitute in equation ( [ eq:470 ] ) for @xmath35 . doing so , and simplifying , we find that @xmath357 } + m_{v,0}\ ] ] which is the limiting magnitude around which an all - sky survey will detect planets . | we develop a method for predicting the yield of transiting planets from a photometric survey given the parameters of the survey ( nights observed , bandpass , exposure time , telescope aperture , locations of the target fields , observational conditions , and detector characteristics ) , as well as the underlying planet properties ( frequency , period and radius distributions ) . using our updated understanding of transit surveys provided by the experiences of the survey teams ,
we account for those factors that have proven to have the greatest effect on the survey yields .
specifically , we include the effects of the surveys window functions , adopt revised estimates of the giant planet frequency , account for the number and distribution of main - sequence stars in the survey fields , and include the effects of galactic structure and interstellar extinction .
we approximate the detectability of a planetary transit using a signal - to - noise ratio ( s / n ) formulation .
we argue that our choice of detection criterion is the most uncertain input to our predictions , and has the largest effect on the resulting planet yield . thus drawing robust inferences about the frequency of planets from transit surveys
will require that the survey teams impose and report objective , systematic , and quantifiable detection criteria . nevertheless , with reasonable choices for the minimum s / n , we calculate yields that are generally lower , more accurate , and more realistic than previous predictions . as examples ,
we apply our method to the trans - atlantic exoplanet survey , the xo survey , and the _ kepler _ mission .
we discuss red noise and its possible effects on planetary detections .
we conclude with estimates of the expected detection rates for future wide - angle synoptic surveys . |
in recent years , quantum turbulence has become an established research area in low - temperature physics.@xcite the similarities and differences of quantum turbulence in relation to classical turbulence comprise one area of interest of quantum turbulence .
for example , although both classical and quantum turbulences consist of eddies or vortices , these two kinds of turbulences are quite different in terms of their hydrodynamic properties .
first , classical fluid can have arbitrary vorticity , while in contrast the vorticity in superfluid @xmath0he is quantized ; circulation around a quantized vortex is @xmath1 , where @xmath2 is planck s constant and @xmath3 is mass of @xmath0he atom , with the core size of a quantized vortex given by the healing length of superfluid @xmath0he , which is about the atomic size of @xmath0he .
moreover , quantized vortices can not be generated under usual experimental conditions , unlike eddies in classical fluids . instead , remnant vortices are essential for the generation of quantum turbulence .
it is believed that remnant vortices are usually pinned to the roughness of the wall.@xcite indeed , hashimoto _ _ et al.__@xcite experimentally showed that quantum turbulence can not be generated without remnant vortices .
second , classical fluid is viscous , while superfluid @xmath0he consists of viscous normal fluid and an inviscid superfluid according to a two - fluid model .
the two fluids are coupled in turbulence through mutual friction , which is the interaction between vortices and normal fluid .
the ratio of the two fluids is temperature dependent , and the normal fluid dominates the fluid if the temperature is close to the transition temperature , while in the limit of zero temperature , the superfluid component dominates the fluid , and pure quantum turbulence is achieved . the attention of the present study is confined to this latter limiting case . in spite of the above distinct properties
, quantum and classical turbulences also exhibit similarities .
kolmogorov s law , which is a statistical law found in a classical turbulence , has also been observed in quantum turbulence both experimentally and numerically.@xcite this law can thus be taken to be universal over quantum and classical turbulences
. under this new motivation , quantum turbulence has recently been studied , and accordingly the progress of experimental techniques has given rise to a variety of methods to create and observe quantum turbulence . to both create and detect quantum turbulence , objects oscillating in a superfluid , such as spheres , grids , thin wires , and tuning forks are often employed.@xcite the most commonly measured physical quantity is the velocity of the oscillating objects . at
temperatures lower than 1 k , the velocity exhibits the clear transition from laminar to turbulent state when the driving force is increased.@xcite for magnitudes of velocities less than about 50 mm / s , the drag force is proportional to the oscillation velocity due to collisions between the oscillating object and some remaining excitations . at higher magnitudes of velocity ,
the drag force is proportional to the square of the oscillation velocity magnitude because turbulence is generated around the object .
these results also describe a similarity between quantum and classical turbulence ; in classical fluid , the drag force on an object in a uniform flow at high reynolds number is described by @xmath4 where @xmath5 is the drag coefficient , @xmath6 is the density of the fluid , @xmath7 is the projection area of the object normal to the flow , and @xmath8 is the flow velocity.@xcite this relation could be applicable to quantum turbulence created by an oscillating object . at low reynolds number , stokes s drag force acts on the object , which is proportional to the magnitude of the flow velocity , with the result that the drag coefficient @xmath5 becomes inversely proportional to the magnitude of the flow velocity .
fitting experimental results to a relationship of the form of eq .
( [ eqn : drag_force ] ) , we derive a constant drag coefficient , which is of order unity , as for classical turbulence . in this paper
, we shall further add to knowledge of the similarities of quantum and classical turbulences by presenting numerical simulations of the drag force and coefficient on the oscillating object .
however , in calculating the drag force numerically , the pressure on the object needs to be calculated over the entire surface of the object .
the pressure at each point on the object can be calculated from the velocity field by euler s equation , and the velocity field is generated by the vortices .
it is possible to calculate the pressure if the number of the vortices is relatively small,@xcite but it is difficult with current computational resources to integrate the pressure in quantum turbulence in cases where a large number of vortices exist . as a consequence , we shall introduce a method to evaluate the drag force .
we assume that the oscillating object does work on the fluid when the vortices are grown by the object and that a drag force is caused as a result of the work .
it is shown later that the drag force can be estimated from the energy given to the fluid by the object in an equilibrium state of quantum turbulence .
the increase in the kinetic energy of the fluid can be evaluated from the energy of vortices grown by the object .
finally , from the energy given per unit time divided by the velocity of the object we derive the drag force . in this paper ,
the drag force in quantum turbulence is targeted for the following reason : the drag force becomes definite according to the method of this paper only if the vortices are grown by the oscillating object . in the laminar regime , it is expected that the drag force is caused by collisions between the object and excitations and that the growth of vortices is not essential . in the turbulent regime , on the other hand
, the drag force is caused by the growth of vortices .
for this reason , we focus on the dynamics of vortices around an oscillating object and solve the equations of motion for the vortices numerically . in the simulation , we consider an object oscillating with a constant magnitude of velocity @xmath9 , and estimate the drag force due to the vortices , that is the drag force @xmath10 is described in the form @xmath11 .
note that this is in contrast to experimental studies , which measure the oscillation velocity of the object for a certain driving force , obtaining the velocity @xmath12 .
the functional relation @xmath11 used in the present simulations is thus inverse to the relation used in the experiments , but it is worth comparing the results for the following reason . as we shall see in sec .
[ turbulence ] , when the object oscillates , remnant vortices are stretched and the vortex tangle grows , which results in the drag force on the object . finally , the vortex tangle reaches an equilibrium state of quantum turbulence .
the drag force is then uniquely determined by the oscillation velocity in the simulation . in an experimental equilibrium state@xcite on the other hand , the drag force responsible for the dissipation by vortices must be equal to the driving force that injects the energy into the system . in addition ,
the oscillation velocity is measured under a constant driving force , namely , drag force in an equilibrium state of turbulence in the experiments .
hence there is one - to - one correspondence between the oscillation velocity and the drag force , which allows us to compare the drag force of the simulation with that of the experiments .
the coupled dynamics of vortices and the oscillating object is not considered in this work , which is validated by verifying that the drag force depends only on the oscillation velocity .
the oscillation of the object causes the growth of vortices , and the rate of the growth depends on the oscillation velocity because the vortices attached to the object are stretched by the object ( details in sec .
[ turbulence ] ) .
the vortex dynamics are influenced only by the magnitude of the oscillation velocity irrespective of whether the dynamics of the object are considered or not .
this is because since the drag force is evaluated from the rate of the vortex growth , the drag force does not depend on the details of the motion of the object but the oscillation velocity .
a previous numerical simulation of an oscillating sphere was performed by hnninen _ _ et al.__@xcite they used the same size of the sphere and frequency of the oscillation as the experiment of shoepe _ _ et al.__@xcite the simulation seems to show a growth of vortices toward turbulence , but an equilibrium state of quantum turbulence is not achieved , which differs from the present results .
this is because in hnninen _
et al._s@xcite study the vortices extending between the wall and the sphere remained attached and continued to generate vortex rings , while in our simulations , mature vortices are soon detached from the object and the remaining vortices are able to grow successively ( details in sec .
[ turbulence ] ) .
it is important to note that the drag force can not be calculated from the simulations of hnninen _ _ et al.__@xcite in the manner of the present study , since our method assumes an equilibrium state .
the contents of this paper are as follows . in sec .
[ formulation ] , we shall clarify the configuration and formulation of the model and introduce the equations of motion required in the dynamics of quantum turbulence .
in sec . [ turbulence ] , we shall describe the process whereby quantum turbulence is generated by an oscillating object . in sec .
[ drag force ] , we shall introduce the present proposed method of evaluating the drag force on the oscillating object and show the dependency of the drag force on the velocity of the object .
section [ conclusion ] is devoted to the conclusion and the discussion .
superfluid @xmath0he at 0 k can be treated mathematically as an ideal incompressible fluid of vanishing viscosity . in a superfluid ,
any circulation is quantized in units of @xmath13 .
the vorticity is infinitely distributed on thin vortex core , whose size is the order of 1 , so that the circulation around the vortex filament is restricted to @xmath13.@xcite with these features , it is valid to adopt the vortex filament approximation , which regards a vortex as a line .
numerical study of quantum turbulence can be accomplished by solving the equations of motion for the vortex filaments that form the turbulence.@xcite according to helmholtz s theorem , a vortex filament moves with superfluid velocity on the vortex . in an infinite system of superfluid , the superfluid velocity is equal to the velocity generated by vortex filaments , which takes the form of the biot - savart integration , @xmath14 where @xmath15 is a position vector on a vortex and the integration is performed over all the vortex lines . in attempting to determine a velocity field at a certain point @xmath16 on a vortex , eq .
( [ eqn : biot - savart ] ) diverges as @xmath17 .
the divergence is evaded by introducing cutoff parameter @xmath18 in eq .
( [ eqn : biot - savart ] ) . in the presence of a boundary
, additional superflow @xmath19 appears so as to satisfy the boundary condition for an inviscid fluid , @xmath20 where @xmath21 is a vector normal to the boundary . while many possible geometries may be considered for the vibrating objects , in the present study we restrict attention to the sphere for simplicity . for a velocity field generated by an infinitesimal element @xmath22 of a closed - loop vortex , eq .
( [ boundary ] ) is satisfied by considering an image vortex inside the sphere.@xcite
the image vortex consists of two parts : a radial image @xmath23 and a tangential image @xmath24 . in fig .
[ fig : img_vor ] , the origin is set at the center of the sphere , @xmath25 is the radius of the sphere , @xmath26 is a radius vector of the vortex element , @xmath27 is a radial component of @xmath22 , and @xmath28 is a tangential component defined as @xmath29 .
the tangential image is located at @xmath30 from the origin and anti - parallel to @xmath28 .
its length is @xmath31 , and it has circulation @xmath32 .
the radial image directs from @xmath33 to the origin , its length is @xmath30 and it has circulation @xmath34 . when the sphere moves with the velocity @xmath35 , the boundary condition [ eq . ( [ boundary ] ) ] is satisfied in a reference frame that moves with the sphere : @xmath36 the extra term in the parenthesis is canceled out by adding the following term to @xmath19 : @xmath37 where @xmath38 is a flow induced by the motion of the sphere and determined by the instantaneous velocity of the sphere.@xcite thus @xmath19 is obtained by adding the contribution from the image vortices and the velocity field given by eq .
( [ eqn : uniform_flow ] ) .
the dynamics of the vortices are completely determined by the above formulation .
when two vortices cross or when a vortex becomes close to a spherical boundary , it is theoretically known that reconnection , which is topological change , occurs.@xcite in the framework of the vortex filament model , the details of the vortex core are neglected , and so these formulations do not describe the dynamics of the reconnection . in the present numerical simulation , an exceptional routine
is included in our numerical code to allow reconnection to occur when a vortex approaches another vortex or a spherical boundary within the numerical resolution.@xcite it is thought that quantum turbulence originates from remnant vortices rather than from the intrinsic nucleation of vortices.@xcite thus we inject vortex rings toward the oscillating object to induce quantum turbulence , considering the experiment of the osaka group.@xcite this approach using the injection of vortices requires the consideration of two parameters : the ring size and the time interval in which the vortices are injected .
the ring size can be estimated experimentally by the time of flight for the vortex ring , and we utilize the value of the ring size 1 @xmath39m.@xcite we assume that the time interval is smaller than the period of the oscillation , say , 0.05 ms.@xcite in order to replicate the conditions of ref
. 15 , in the present study the same parameter values as this previous study are used , such as the radius of sphere and the oscillation frequency of the sphere .
the diameter of the sphere is 3 @xmath39 m , the frequency of the oscillation is 1590 hz , while the oscillation velocity is chosen in the range of 3090 mm / s .
[ cols="^,^ " , ] table [ tab : cd ] shows the drag coefficient calculated using different time intervals at an oscillation velocity of magnitude 50 mm / s .
as the injection interval becomes shorter , the rate of vortex growth increases , resulting in an increase in the drag coefficient .
however , it may be noted that the drag coefficient remains of order unity .
quantum turbulence was created around an oscillating sphere , and its evolution was studied .
we proposed a method to evaluate the drag force in quantum turbulence .
the drag force was calculated by considering the energy supplied by the sphere , and it is proportional to the square of the magnitude of the oscillation velocity .
this dependency on the velocity is quite similar to that in classical turbulence .
the drag coefficient was also calculated , and its order of magnitude is in agreement with that of the previous experimental results . the similarity between classical and quantum turbulences was confirmed by numerical simulations .
we believe that the drag coefficient has an universal value independent of the details of the system .
our future work is to confirm whether the drag coefficient still remains of the order of unity even if the parameters such as the frequency of the oscillation and the geometry of the oscillating object are changed .
the simulations were performed for a smooth sphere , but the objects used in the experiments have surface roughness .
the roughness may increase the drag force and the drag coefficient ; when a vortex is detached from the sphere due to the oscillation , it is likely to leave a small vortex bridge over the pinning sites on the surface .
the small vortex plays a role of `` seed '' for the growth of the vortex , and this is equivalent to injecting more vortices to the sphere , resulting in an increase in the drag force and the drag coefficient .
however , there is no clear way to proceed to obtain the solution of the equations of motion of vortices under the boundary conditions which describe the surface roughness at present , and the numerical simulations require a model that phenomenologically depicts the vortex pinning .
figure [ fig : drag_force ] enables us to discuss the transition from the turbulent to laminar state .
below oscillation velocities of magnitude 50 mm / s , the drag force starts to deviate from eq .
( [ eqn : drag_force ] ) and tends to zero .
this is because the motion of the sphere is so slow that vortices can not be stretched before the attached vortices leave the sphere , resulting in the drag force tending to zero .
this means that the oscillating sphere can no longer sustain turbulence , and that the flow regime will return to a laminar state . from dimensional analysis , the critical velocity
at which turbulence returns to laminar flow is given by @xmath40 , where @xmath41 is constant with the order of unity , and this formulation agrees with the critical velocity in the experiments.@xcite in the simulation , @xmath42 m@xmath43/s , @xmath44 rad / s , and finally one obtains @xmath45 30 mm / s .
this is very close to the velocity 50 mm / s below which the deviation from eq .
( [ eqn : drag_force ] ) starts in the simulation .
therefore the velocity magnitude 50 mm / s can be considered in the simulations to represent the critical velocity magnitude .
is very grateful for helpful discussion with risto hnninen .
m.t . acknowledges the supports of grant - in - aid for scientific research from jsps ( grant no .
18340109 ) and grant - in - aid for scientific research on priority areas ( grant no .
17071008 ) from mext .
30 m. tsubota , j. phys .
77 * , 111006 ( 2008 ) .
_ progress in low temperature physics _ , edited by w. p. halperin and m. tsubota ( north - holland , amsterdam , 2008 ) , vol . 16 .
r. j. donnelly , _ quantized vortices in helium ii _
( cambridge university press , cambridge , 1991 ) . n. hashimoto , r. goto , h. yano , k. obara , o. ishikawa , and t. hata , phys . rev .
b * 76 * , 020504(r ) ( 2007 ) . c. nore
, m. abid , and m. e. brachet , phys .
lett . * 78 * , 3896 ( 1997 ) .
j. maurer and p. tabeling , europhys .
* 43 * , 29 ( 1998 ) .
s. r. stalp , l. skrbek , and r. j. donnelly , phys .
lett . * 82 * , 4831 ( 1999 ) .
t. araki , m. tsubota , and s. k. nemirovskii , phys .
lett . * 89 * , 145301 ( 2002 ) .
m. kobayashi and m. tsubota , phys .
lett . * 94 * , 065302 ( 2005 ) j. phys .
74 * , 3248 ( 2005 ) .
j. jger , b. schuderer , and w. schoepe , phys .
lett . * 74 * , 566 ( 1995 ) .
s. n. fisher , a. j. hale , a. m. gunault , and g. r. pickett , phys .
lett . * 86 * , 244 ( 2001 ) .
h. a. nichol , l. skrbek , p. c. hendry , and p. v. e. mcclintock , phys .
lett . * 92 * , 244501 ( 2004 ) .
d. i. bradley , d. o. clubb , s. n. fisher , a. m. gunault , r. p. haley , c. j. matthews , g. r. pickett , v. tsepelin , and k. zaki , phys .
lett . * 95 * , 035302 ( 2005 ) .
h. yano , n. hashimoto , a. handa , m. nakagawa , k. obara , o. ishikawa , and t. hata , phys .
b * 75 * , 012502 ( 2007 ) .
r. goto , s. fujiyama , h. yano , y. nago , n. hashimoto , k. obara , o. ishikawa , m. tsubota , and t. hata , phys .
* 100 * , 045301 ( 2008 ) .
m. blakov , m. loveko , v. b. eltsov , e. gao , r. de graaf , j. j. hosio , m. krusius , d. schmoranzer , w. schoepe , l. skrbek , p. skyba , r. e. solntsev , and w. f. vinen , j. low temp . phys . * 150 * , 525 ( 2008 ) .
l. skrbek and w. f. vinen , in _ progress in low temperature physics _ , edited by w. p. halperin and m. tsubota ( north - holland , amsterdam , 2008 ) , vol .
16 , p. 193 .
d. kivotides , c. f. barenghi , and y. a. sergeev , phys .
b * 75 * , 212502 ( 2007 ) * 77 * , 014527 ( 2008 ) .
r. hnninen , m. tsubota , and w. f. vinen , phys .
b * 75 * , 064502 ( 2007 )
. k. w. schwarz , phys .
b * 31 * , 5782 ( 1985 ) . k. w. schwarz , phys . rev .
b * 38 * , 2398 ( 1988 ) .
p. g. saffman , _ vortex dynamics _ ( cambridge university press , cambridge , 1992 ) .
an alternative solution can also be used and is valid even for a vortex ending at boundaries that extend infinitely .
however , the method of the image vortex reduces the required computational time and is advantageous for this calculation .
k. w. schwarz , phys .
rev . a * 10 * , 2306 ( 1974 ) .
j. koplik and h. levine , phys .
* 71 * , 1375 ( 1993 ) .
m. tsubota , t. araki , and s. k. nemirovskii , phys .
b * 62 * , 11751 ( 2000 ) . see epaps document no .
e - prmbdo-79 - 010909 for the movie of turbulence generation for the case of a sphere oscillating with a velocity magnitude of 90 mm / s . for more information on epaps ,
see http://www.aip.org/pubservs/epaps.html l. d. landau and e. m. lifschitz , _ fluid mechanics_,2nd ed .
( pergamon , london , 1987 ) .
note that the computational box is different from @xmath46 .
the volume @xmath46 is considered large enough so that the vortices never escape from @xmath46 in the time period @xmath47 , while vortices can escape from the computational box within @xmath47 . in the time interval
@xmath47= 100 ms , many vortices escape from the computational box as easily found in fig .
[ fig : length ] , and the statistical fluctuation of @xmath48 is supposed to be small , we have not yet estimated the fluctuation of @xmath48 systematically though .
r. hnninen and w. schoepe , j. low temp . phys . * 153 * , 189 ( 2008 ) . | this paper reports results of the computation of the drag force exerted on an oscillating object in quantum turbulence in superfluid @xmath0he .
the drag force is calculated on the basis of numerical simulations of quantum turbulent flow about the object .
the drag force is proportional to the square of the magnitude of the oscillation velocity , which is similar to that in classical turbulence at high reynolds number .
the drag coefficient is also calculated , and its value is found to be of the same order as that observed in previous experiments . the correspondence between quantum and classical turbulences is further clarified by examining the turbulence created by oscillating objects . |
little is known about the properties of the dense gas component on gmc scales in external galaxies yet it is dense gas ( @xmath5 @xmath6 ) that most directly correlates with global star formation rates ( e.g. , * ? ? ?
* ; * ? ? ?
* ) . to understand the evolution and regulation of current star formation in galaxies , dense gas properties need to be accurately characterized . in a recent imaging survey of millimeter - wave molecular lines in the nearby spiral nucleus of ic 342 , a surprising degree of morphological variation
is observed in its dense gas @xcite , which was not anticipated from co or hcn ( e.g. , * ? ? ?
* ; * ? ? ?
these and other observations ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) confirm that co and hcn fail to fully constrain excitation and chemical properties of the dense clouds .
various probes of the dense component exist , each with various strengths and weaknesses ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
hcn(10 ) is the most commonly used dense gas tracer , primarily because it is bright , however , it is optically thick in most starburst environments ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , has a widely spaced rotational ladder with its j@xmath73 transitions in the submm and can be sensitive to chemical effects ( pdr effects and ir pumping ; e.g. * ? ? ?
* ; * ? ? ?
* ) . here
we present a study of the physical conditions of the very dense gas component of a nearby starburst nucleus using hc@xmath0n . with its large electric dipole moment ( @xmath8 = 3.72 debye versus 3.0 debye for hcn ) , low opacity and closely spaced rotational ladder accessible to powerful centimeter and millimeter - wave interferometers , hc@xmath0n is well - suited to high resolution imaging of the structure and excitation of the densest component of the molecular gas ( eg .
* ; * ? ? ?
* ; * ? ? ?
hc@xmath9n(54 ) has a critical density of @xmath10 @xmath6 and an upper level energy of e / k = 6.6 k ; the 1615 line has a critical density of @xmath11 @xmath6 and an upper level energy of e / k = 59 k. we have observed the j = 54 and j = 1615 lines of hc@xmath0n at @xmath122 resolution in ic 342 with the very large array ( vla ) and the plateau de bure interferometer ( pdb ) , respectively .
ic 342 is one of the closest ( d@xmath133 mpc , or 2 = 29 pc ; eg . * ? ? ?
* ; * ? ? ?
* ) , large spirals with active nuclear star formation @xcite .
it is nearly face - on , with bright molecular line emission both in dense clouds and a diffuse medium @xcite .
the wealth of data and proximity of ic 342 permit the connection of cloud properties with star formation at sub - gmc spatial scales .
the data published here represent the first high resolution maps of hc@xmath9n(54 ) and hc@xmath9n(1615 ) in an external galaxy .
excitation is a key component in the interpretation of molecular line intensities .
so these two maps are compared with the previously published lower resolution hc@xmath0n(109 ) map made with the owens valley millimeter array @xcite , to constrain the physical conditions of the densest component of the ism in the center of ic342 , and to correlate properties of the dense gas with star formation , diffuse gas , and chemistry .
aperture synthesis observations of the hc@xmath0n j=54 rotational line ( 45.490316 ghz ) towards ic 342 were made with the d configuration of the vla on 2005 november 25 ( vla i d : am839 ) .
the synthesized beam is 195@xmath1415 ( fwhm ) ; position angle ( pa)=-11.2 .
fifteen 1.5625 mhz channels were used , for a velocity resolution of 10.3 km s@xmath15 , centered at @xmath16 km s@xmath15 .
the phase center is @xmath17(j2000 ) = 03@xmath1846@xmath19483 ; @xmath20(j2000 ) = 6805470 . amplitude tracking and pointing was done by observing the quasar 0228 + 673 every 45 minutes .
phases were tracked by fast switching between the source and the quasar 0304 + 655 with 130s/50s cycles .
absolute flux calibration was done using 3c48 and 3c147 and is good to @xmath13510% .
calibration and analysis was performed with the nrao astronomical image processing software ( aips ) package .
the naturally weighted and cleaned datacube has an rms of 0.70 mjy beam@xmath15 .
correction for the primary beam ( @xmath131 at 45 ghz ) has not been applied .
the shortest baselines in the dataset are @xmath215 k@xmath22 , corresponding to scales of @xmath1340 ; structures larger than this are not well - sampled .
the hc@xmath0n(16 - 15 ) line emission in the center of ic 342 was observed with 5 antennas using the new 2 mm receivers of the iram plateau de bure interferometer ( pdbi ) on 2007 december 28 and 31 in c configuration with baselines ranging from 24 to 176 m .
the phase center of the observations was set to @xmath17(j2000 ) = 03@xmath2346@xmath1948105 ; @xmath20(j2000 ) = 68054784 . nrao150 and 0212 + 735 served as phase calibrators and were observed every 20 min .
flux calibrators were 3c84 and 3c454.3 .
the calibration was done in the gildas package following standard procedures .
the hc@xmath0n(16 - 15 ) line at 145.560951ghz was observed assuming a systemic velocity of @xmath24 km s@xmath15 and a spectral resolution of 2.5mhz ( 5.15 km s@xmath15 ) .
the average 2 mm continuum was obtained by averaging line - free channels blue and redward of the hc@xmath0n and h@xmath4co lines and subtracted from the @xmath25 datacube to obtain a continuum - free datacube .
the final naturally weighted and cleaned datacube with 10kms@xmath15 wide channels has a clean beam of [email protected] , pa=46@xmath27 and an rms of 1.7 mjy / beam .
single - dish observations of hc@xmath0n(1615 ) find a peak brightness temperature of 6.7 mk in a 16.9@xmath28 beam @xcite .
convolving our map to this beam and sampling it at the same location yields a peak brightness of 7.1 mk , which agrees within the uncertainties of both datasets .
therefore no flux is resolved out of the interferometer maps , as expected for these high density tracers .
continuum subtracted integrated intensity line maps of hc@xmath0n(54 ) , hc@xmath0n(109 ) and hc@xmath0n(1615 ) in ic 342 are shown in figure [ inti ] .
figure [ inticont ] shows the 54 transition overlaid on the 7 mm continuum image generated from offline channels .
locations of giant molecular cloud ( gmc ) cores @xcite and the optical clusters ( e.g. * ? ? ?
* ) , compared with hc@xmath0n(54 ) and hcn(10 ) @xcite , are also shown in figure [ inticont ] .
figure [ spec ] shows the hc@xmath9n(54 ) and hc@xmath0n(16 - 15 ) spectra ( in flux units ) taken over the same 2aperture centered on each cloud .
hc@xmath0n(54 ) emission picks out most clouds seen in other dense gas tracers ( e.g hcn(10 ) .
the only labeled cloud not clearly detected in hc@xmath9n(54 ) is gmc b , the cloud associated with the nuclear star - forming region ( @xmath29 ; * ? ? ?
* ; * ? ? ?
* ) . positions of the gmcs measured in hc@xmath0n are consistent with those fitted in c@xmath1o(2 - 1 ) to within a beam @xcite .
an additional gmc is detected in hc@xmath0n(54 ) just south of gmc c3 , labeled c4 . unlike c@xmath1o(2 - 1 ) which peaks at gmc c2 , hc@xmath0n(54 ) emission peaks farther north , towards gmc c1 , at a distance of 105 pc from the nucleus suggesting changes in excitation across gmc
c. gmcs a and d are resolved into two components , but we do not discuss them as separate entities given the lower signal - to - noise , other than to state that there is a clear difference in velocity centroid between the 54 and 1615 transitions ( e.g. figure [ spec ] ) indicating there is likely an excitation gradient across gmc a. no emission @xmath30 is detected towards the weak co(2 - 1 ) feature associated with the central nuclear star cluster @xcite . the j=1615 line intensity is comparable to that of j=54 line in all clouds . while dominated by gmc c , hc@xmath0n(16 - 15 ) is detected ( or tentatively detected ) towards all clouds , except gmc d. fluxes for gmcs d and d are quite uncertain since they are just inside the half power point of the pdbi primary beam .
that gmc c is much brighter than the other clouds in both transitions indicates large quantities of dense gas are present in this cloud .
hc@xmath0n(1615 ) favors c2 over c1 .
gmc b , while still faint in an absolute sense , is significantly brighter in hc@xmath0n(1615 ) relative to 54 .
this is not unexpected given the higher excitation requirements of the 1615 line and that both b and c2 have 7 mm continuum sources associated with massive star formation .
comparisons of the hc@xmath9n ( 54 ) , ( 109 ) and ( 1615 ) maps provide a chance to establish the excitation of the densest molecular cloud gas component .
line intensities from the j=54 and j=16 - 15 lines were measured over 2 apertures centered on each of the gmcs .
table [ tinti ] records gaussian fits to each spectrum along with integrated line intensities .
peak antenna temperature ratios are calculated for hc@xmath0n(1615)/hc@xmath0n(54 ) , hereafter denoted r@xmath32 .
r@xmath32 ranges from less than 0.06 up to @xmath130.5 , with gmc b having the highest value ( table [ trat ] ) .
the 54 and 1615 transitions generally bracket the peak of the level populations , so we achieve good constraints on gas excitation .
lte excitation temperatures , t@xmath33 , implied by r@xmath32 range from @xmath1210 k to @xmath718 k ( where we have neglected t@xmath34 in this determination ) .
excitation is lowest towards gmcs d , d and e. these excitation temperatures t@xmath33 are similar to those found from the presumably much less dense gas traced in c@xmath1o ( table 2 ; * ? ? ?
* ) . only the gmc c clouds have significantly higher t@xmath33 in hc@xmath9n towards c1 - c3 t@xmath33(hc@xmath0n ) are about a factor of two greater than t@xmath33(c@xmath1o ) .
the resolution of the hc@xmath0n(109 ) data is significantly lower than it is for the 54 and 1615 lines . to compare j=54 and j=10 - 9 line intensities ,
the 54 data were convolved to the resolution of the ( 109 ) data ( @xmath35 * ? ? ?
* ) , then integrated intensity ratios , hereafter r@xmath36 , were sampled at the locations of r@xmath32 .
though at lower resolution than r@xmath32 , we make the approximation that r@xmath36 does not change on these sub - gmc scales .
while leading to larger uncertainties , this provides a way to include all three transitions in modeling dense gas excitation at very high resolution .
@xmath37 range from 0.25 to 1.4 ( table [ tinti ] ) .
gmc b has the highest value of 1.4 .
the remainder of the gmcs have ratios of @xmath38 .
excitation temperatures implied by these ratios range from t@xmath39 616 k , consistent with those derived from @xmath40 separately .
t@xmath33 derived from r@xmath36 are similar to those derived from c@xmath1o .
the only exception here is gmc a , the cloud where pdrs ( photon - dominated regions ) dominate @xcite .
towards gmc a t@xmath33(c@xmath1o ) and t@xmath33(r@xmath32 ) are 12 - 13 k , twice that of t@xmath33(r@xmath36 ) . before modeling the densities implied by the line ratios
, we test whether ir pumping can be responsible for the observed excitation .
ir pumping is important if , @xmath41 where @xmath42 is the einstein @xmath43 of the corresponding @xmath44 vibrational transitions at @xmath45 m , @xmath46(45 @xmath8 m ) is the 45 @xmath8 m intensity as seen by the hc@xmath0n molecules and @xmath47 is the einstein @xmath48 of the rotational state ( e.g. * ? ? ?
it is extremely difficult to estimate the applicable 45 @xmath8 m ir intensity , but it is expected to be most intense towards the starburst gmc ( b ) .
a detailed assessment of ir pumping must await high resolution mir maps , but we constrain @xmath46(45 @xmath49 ) in several ways .
first , we take the 45 @xmath8 m flux from @xcite and scale it by the fraction of total 20 @xmath8 m flux that comes from within 2.1@xmath28 of the starburst as found by @xcite and then average over that aperture . for the average @xmath46(45 @xmath8 m ) calculated this way ,
@xmath50(45 @xmath8 m ) is @xmath51 ( @xmath52 ) times too low to pump the 1615 ( 54 ) transition . alternatively
if we ( very conservatively ) take the total mir luminosity from the central 30@xmath28 and assume it comes from a blackbody of the observed color temperature ( e.g. @xmath1350 k ; * ? ?
* ) then @xmath46(45 @xmath8 m ) is still at least an order of magnitude too low to meet the inequality in eq .
[ pump ] for both transition .
ir pumping rates only become comparable to @xmath47 for the 54 transition if the total ir luminosity originates from a @xmath132.5 pc source with a source temperature of @xmath53100 k. we conclude that ir pumping is not important for the 1615 transition of hc@xmath0n in any reasonable geometry of the ir field . for the 54 transition to be sensitive to ir pumping ,
the ir source must be warm , opaque and extremely compact .
therefore ir pumping is neglected for all clouds in the following discussion .
the values of the excitation temperature constrain the density and kinetic temperatures , @xmath54 and @xmath55 , and the physical conditions of the clouds driving the excitation . a series of large velocity gradient ( lvg )
radiative transfer models were run to predict the observed intensities and line ratios for a given @xmath54 , t@xmath56 and filling factor , @xmath57 , of the dense component ( e.g. , * ? ? ?
single component lvg models are instructive , particularly when lines are optically thin , as is the case for hc@xmath0n .
the lvg model used is that of @xcite , adapted to hc@xmath0n , with levels up to j=20 included .
collision coefficients are from @xcite .
a range of densities , @xmath54 = 10@xmath5810@xmath59 , and kinetic temperatures , t@xmath56 = 0100 k , was explored .
column densities based on lte excitation ( table [ trat ] ) , are calculated at 2@xmath28 resolution from : @xmath60 using molecular data of @xcite , hc@xmath0n(54 ) intensities , and t@xmath33(hc@xmath0n ) from table [ trat ] .
hc@xmath0n abundances are found to be , @xmath61 , with the highest values towards c1 and d. while uncertain these abundances agree with those found in @xcite and are typical of galactic center hc@xmath0n abundances ( e.g. * ? ? ?
* ; * ? ? ?
* ) and good enough for constraining @xmath62
. the ratio of cloud linewidth to core size is @xmath63 for the gmcs .
therefore , we adopt a standard model abundance per velocity gradient of @xmath62 = 10@xmath64 km s@xmath15 , but run models for values of @xmath65 - @xmath66 .
antenna temperatures are sensitive to the unknown filling factor . in these extragalactic observations
, the beam corresponds to scales large compared to cloud structure , and hence filling factors are not directly known .
to first order , the line ratios , @xmath37 and r@xmath32 are independent of filling factor if we assume that hc@xmath0n(54 ) , ( 109 ) and ( 1615 ) originate in the same gas .
so model ratios are compared to the observed data to constrain parameter space . for the parameter space implied by the line ratios
model brightness temperatures are determined .
a comparison of the model brightness temperature to the observed brightness temperature sets the required areal filling factors for that solution .
figure [ lvg ] displays the results of the lvg modeling .
acceptable @xmath67 ( t@xmath56,@xmath54 ) parameter spaces are shown for each line ratio ( r@xmath36 , thick solid gray line ; r@xmath32 , thick solid black line ) .
also shown in figure [ lvg ] are the acceptable solutions obtained from the c@xmath1o(21)/c@xmath1o(10 ) line ratio ( thin gray lines ; * ? ? ?
figure [ levels ] and table [ tlvg1 ] display the model flux versus upper j state of the line for the t@xmath68 = 10 k , 30 k and 50 k solutions ( also 70 k solutions for gmc b ) .
hc@xmath0n(54 ) has a critical density of @xmath69 @xmath6 and an upper energy state of 6.55 k. hc@xmath0n(1615 ) has a critical density , @xmath70 @xmath6 and an upper energy state of 59 k. hc@xmath0n(109 ) values are intermediate .
therefore @xmath37 and r@xmath32 are sensitive probes of @xmath54 between the range of @xmath71 and @xmath72 @xmath6 . observed ratios constrain @xmath54 to @xmath730.1 dex at a given t@xmath56 and to @xmath730.3 dex for all modeled t@xmath74 k. for t@xmath75 40 k , ratios are largely insensitive to t@xmath56 ( curves are horizontal ) .
the lvg models do not constrain density when t@xmath56 is low . thus to narrow the range of possible solutions requires two or more line ratios , or an external constraint on one of the axes .
kinetic temperatures of the dense component at 2@xmath28 scales in ic 342 remain largely unconstrained to date
. one method of constraining the kinetic temperature is the peak t@xmath76 at this resolution of the optically thick co(2 - 1 ) line . using this method ,
t@xmath77 3545 k towards gmcs a , b and c , and t@xmath77 1520 k for gmcs d , d and e @xcite .
these values are not highly discordant from arcminute resolution t@xmath78 k measurements from nh@xmath0 @xcite or the far - infrared dust temperature of 42 - 55 k @xcite .
however , the co(2 - 1 ) traces gas with densities two to three orders of magnitude lower than these hc@xmath0n transitions , so it is not clear that this is the relevant t@xmath56 for the dense cores of the gmcs .
valid solutions are found for the gmcs , with the exception of gmc d , which is the most uncertain due to its distance from the center of the field .
( no fit is attempted for gmc d because both hc@xmath0n(54 ) and hc@xmath0n(1615 ) are upper limits or tentative detections . )
the solutions show modest cloud excitations given the strong nuclear star formation .
good agreement is observed between solutions found using r@xmath32 and r@xmath36 when kinetic temperatures are low ( t@xmath79 k ) .
given uncertainties in line intensities and @xmath80 , higher temperatures can not be ruled out . while t@xmath56 itself is less
well constrained the combination @xmath54t@xmath56 is well determined .
since line brightnesses are the measured quantity , as gas temperatures are raised , densities or filling factors must decrease to compensate .
the nature of the solutions are such that the densities decrease more than the filling factors ( table [ lvg ] ) .
gmcs a , c1 and c3 are best fit with @xmath81 @xmath6 and t@xmath77 20 k. for t@xmath82 k derived densities drop to @xmath83 @xmath6 , while they drop to @xmath84 @xmath6 for t@xmath56 = 100 k. changes in @xmath62 do not strongly influence the derived solutions . towards these clouds filling factors
are @xmath850.03 - 0.1 .
for comparison , a 1 pc@xmath58 cloud would have @xmath86 0.0012 for the aperture size and ic 342 s distance .
therefore these gmcs appear to have a few dozen dense clumps similar to the larger clumps found in galactic gmcs ( e.g. * ? ? ?
* ; * ? ? ?
gmc c2 appears to have similar densities but slightly elevated temperatures , t@xmath77 30k . in gmcs d and d
densities are at least a factor of four lower than the other clouds .
little can be said about the kinetic temperatures of these gmcs .
gmc d is the only location with statistically discrepant solutions from r@xmath36 and r@xmath32 . here
the ( 1615 ) line is somewhat brighter than would be expected from the r@xmath36 .
the starburst gmc , b , is markedly different from the others .
the combination of a low absolute hc@xmath0n brightness temperature and high ratios , @xmath37 and r@xmath32 , require higher densities of @xmath87 @xmath6 and a smaller ( @xmath880.02 ) filling factor .
both line ratios seem to indicate a kinetic temperature of more than t@xmath77 40 k , and perhaps significantly higher , averaged over our 30 pc beam .
the dense gas in gmc b , which is closest to the active star - forming region and ir source , is more compact , denser , and more highly excited than the other clouds . for our own galactic center
it has long been argued that molecular clouds must be unusually dense , at least @xmath89 to withstand tidal forces ( eg . * ? ? ? * ) .
if one assumes that gmc c , the most well defined cloud , is a point mass in a spherical mass potential of the galaxy , its roche limit would be at a radius of 25 pc . the radius of the combined c1-c2-c3 complex is @xmath725 pc .
in ic 342 , the proximity to the nucleus where the strong noncircular velocity field @xcite together with feedback from the starburst @xcite suggests that these clumps have densities large enough to maintain their identity but likely will not remain gravitationally bound to each other . for all gmcs ( including gmc b ) ,
t@xmath33(hc@xmath0n ) @xmath21 t@xmath33(c@xmath1o ) .
if a model in which all the h@xmath4 exists in one uniform component , then the similarity of t@xmath33 from both hc@xmath0n and c@xmath1o implies that the densities of the molecular clouds are high enough ( @xmath90 @xmath6 ) to thermalize both c@xmath1o and hc@xmath0n across the nucleus .
( lvg solutions for c@xmath1o j=21 and 10 lines from @xcite are shown with the hc@xmath9n solutions in figure [ lvg ] . ) in this monolithic model the low observed t@xmath33 of 1020 k demand that the thermalized clouds must be quite cool .
these temperatures are similar to those of dark clouds in the disk of our galaxy , which is somewhat surprising given the elevated star formation rate in the nucleus of ic 342 .
however this model of a monolithic , dense , cold ism seen in both hc@xmath0n and c@xmath1o runs into problems with total mass constraints . if the ism is uniformly this dense and cold then the total dense gas cloud mass implied by the required densities and filling factors become very large , greater than permitted based on the optically thin c@xmath1o line emission . in table
[ tlvg1 ] the total mass of dense gas , @xmath91 , is approximated from the lvg solutions assuming @xmath92 , with r defined as @xmath93 .
table [ trat ] lists the total molecular gas mass , @xmath94 , estimated from c@xmath1o(2 - 1 ) data ( see * ? ? ?
* ) assuming @xmath95 .
one can see that @xmath91 is typically larger than @xmath94 for t@xmath96 30 k and diverges rapidly as t@xmath56 drops towards the lvg favored values . in short
if gmcs have uniformly such high densities and cold temperatures then the clouds would contain too much h@xmath4 mass for what is observed in c@xmath1o .
to match the line ratios while not violating mass constraints requires a multi - component dense ism . a possible two component model of the dense gas
is discussed in @xmath97 [ cloudsxco ] .
a consideration of the conversion factor between co intensity and h@xmath4 column density , @xmath98 , in the nuclear region of ic 342 suggests that the model of a single , high density and relatively cool ism is not consistent with observations of co isotopologues
. moreover these clouds are unlikely to resemble galactic disk gmcs in their internal structure and dynamics .
the well known galactic relation between co intensity and h@xmath4 column density , @xmath99 can be explained if gmcs consist of optically thick ( in co ) turbulent clumps in virial equilibrium ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
for the clumps to emit in hc@xmath0n densities must be large enough so that t@xmath76 can legitimately be approximated by t@xmath56 .
if we adopt this model , then @xmath98 is approximated by : @xmath100~\rm cm^{-2}\ , ( k~km\,s^{-1})^{-1 } \label{eqxco}\ ] ] where @xmath101 is the volume filling factor and @xmath102 is the density of the _ clumps _ @xcite . from the brightness of the co isotopic lines ,
@xmath103 on 2 scales @xcite .
the conversion factor @xmath98 within the central few hundred pc of ic 342 has been determined to be @xmath104 @xcite .
if the monolithic model presented in the previous section is correct and clouds are virialized clumps in equilibrium , then @xmath105 must be @xmath1060.75 to match the observed conversion factor .
table [ tlvg1 ] records @xmath105 for each of the lvg solutions across the nucleus of ic 342 . for all solutions , @xmath105 is greater than the required value , especially at low t@xmath56 .
for the single component lvg favored t@xmath56 of @xmath1310 - 20 k towards the more quiescent clouds , @xmath107575 , implying @xmath108 should be several times larger than @xmath109 . however
@xmath98 in the nucleus of ic 342 is 34 times _ lower _ than the galactic conversion factor , based on both optically thin isotopologues of co as well as dust emission @xcite .
lower conversion factors appear to be the norm for the nuclei of gas - rich star forming galaxies , including our own ( eg . * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . in the nuclear region of ic 342 dense gas kinetic temperatures do not appear to be higher than the galactic disk value by an amount large enough to offset the observed increase in density over typical galactic disk - like clouds ; t@xmath110100 k would be required .
recently @xcite completed a more generalized treatment of the conversion factor including radiative transfer and concludes that the true exponential dependences of @xmath54 and t@xmath56 are weaker than 0.5 and -1 , respectively .
however in all cases modelled , the dependences remain positive for @xmath54 and negative for t@xmath56 . in summary ,
it is not possible to reconcile bright hc@xmath0n emission from uniformly cool , dense gas ( low t@xmath56 , high @xmath54 and high @xmath111 ) with the known total amount of h@xmath4 present if virialized constant density clumps are adopted .
the lack of an apparent connection between @xmath105 of the dense component and the observed conversion factor indicates that the clouds in the nucleus of ic 342 can not be treated as simple virialized collections of uniform density clumps .
one might expect that a large fraction of the c@xmath1o emission could originate from a moderate density component that is distinct from the denser , hc@xmath0n - emitting gas .
( note that @xmath113co traces a distinct component more diffuse than that traced both by c@xmath1o @xcite and hc@xmath0n . ) as a first order extension to the basic virialized clumps model , we imagine two spatially well - mixed sets of clumps ; one low density c@xmath1o emitting , @xmath114 , and one high density , @xmath115 , that emits both c@xmath1o and hc@xmath0n . assuming the fraction , by number , of clumps with high density is @xmath116 , then if both clumps have the same t@xmath56 the virialized clumps @xmath98 relation becomes : @xmath117~\rm cm^{-2}\ , ( k~km\,s^{-1})^{-1 } , \label{eqxco2}\ ] ] where @xmath118 .
the ratio of dense gas to total h@xmath4 becomes , m@xmath119/m@xmath120 = @xmath121 . for a given t@xmath56
, @xmath114 is chosen to match the c@xmath1o lvg solution @xcite , while @xmath115 is chosen to match the hc@xmath0n lvg solutions , and @xmath122 .
table [ tlvg2 ] displays the adopted c@xmath1o lvg solutions , along with the new @xmath123 and m@xmath119/m@xmath124 for the two component model , assuming , for simplicity , that t@xmath56 is the same for both components .
@xmath123 and hence @xmath98 are decreased from the simple one component model by factors of at least four .
this simple extension results in closer agreement between the observed and calculated values of x@xmath112 ( especially when including the somewhat super - virial line widths ) , while maintaining most of the mass in the dense ( low filling factor ) component .
it is almost certainly the case that this extension is an oversimplification .
in reality , we expect the clouds to exhibit a continuum of densities .
however the lvg modeling demonstrates that at least these three components are required to match the multi - transition observations of @xmath113co , @xmath125co , c@xmath1o and hc@xmath0n .
this is consistent with conclusions from recent single - dish modeling of higher j transitions of hc@xmath0n @xcite , but we find multiple components are required to match intensities not only between gmcs but within individual gmcs .
hcn(10 ) emission is the workhorse for relating quantities of dense gas to star formation ( e.g. * ? ? ?
it is interesting to compare conclusions about the dense component from hc@xmath0n with those of the more commonly used hcn(10 ) .
hcn(10 ) has been mapped at similar spatial resolution by @xcite .
figure [ inticont ] compares hc@xmath0n(54 ) to hcn(10 ) . while hc@xmath0n generally traces the same dense gmcs seen in hcn their relative brightnesses are rather different . in hcn(10 ) gmc a , b and c are all within 10 % of the same brightness and gmcs d and e are 100 % and 50 % weaker , respectively .
whereas in hc@xmath0n(54 ) and ( 1615 ) , c dominates , b is nearly absent and a is not significantly different from d and e. comparisons with hc@xmath0n clearly demonstrate that there is larger variations in dense gas properties than hcn indicates .
the dominate difference is enhanced hcn towards the starburst and gmc a relative to gmc c. unlike the hc@xmath0n transitions , hcn(10 ) is optically thick and has slightly larger ( 25 - 50 % ) filling factors @xcite . as kinetic temperatures increase optically thick transitions brighten more rapidly than ( lower excitation ) optically thin transitions .
hence it is expected that the hcn emission should favor somewhat warmer dense gas .
likely this effect results in the much brighter relative hcn(10 ) intensities towards the starburst . the relative enhancement towards gmc a is less clear .
however , this cloud is known to be dominated by pdrs @xcite , strongly influenced by mechanical feedback from the nuclear cluster @xcite and has a complicated hc@xmath0n temperature and velocity structure ( sections [ morp ] and [ ratio ] ) .
this cloud must have a complex density and temperature structure , potentially with a warmer intermediate density medium ( table [ trat ] ) .
moreover , hcn abundances can be elevated in pdrs .
we conclude that while hcn(10 ) does a good job locating the dense gas but it does a poorer job tracing small scale variation in the properties of the dense gas when a mix of strong star formation and quiescent gas are present .
it is expected that such effects should become more important as specific star formation rates increase and spatial scales decrease .
the brightest hc@xmath0n emission and the brightest star - forming regions do not coincide .
most of the current star formation , traced by bright infrared and thermal radio continuum emission @xcite , is situated about 50 pc to the southwest of the dynamical center , in the vicinity of gmc b. the brightest hc@xmath0n emission by far is on the northeast side of the nucleus , in the northern molecular arm centered at gmc c2 .
this region has free - free emission amounting to only a third of the brightness of the strongest radio source , associated with gmc b. the faintness , in absolute terms , of hc@xmath0n(1615 ) towards gmc b is unexpected and important reflecting a much lower areal filling of warm , very dense gas here .
either the number of cloud clumps or their size is small relative to the other gmcs . on the other hand
, the excitation of hc@xmath0n does reflect the presence of young forming stars .
hc@xmath0n(1615 ) and hc@xmath0n(109 ) are relatively brighter towards the current starburst .
this suggests that the absence of a correlation between hc@xmath0n(54 ) intensity and star formation is partly due to depopulation of the lower energy transitions .
gmcs that are faint in the hc@xmath9n j=1615 transition and not associated with strong star formation show up well at the lower j transitions of hc@xmath0n , as would be expected for gas of lower excitation .
clearly there is a large amount of dense gas currently not actively forming stars , that shows up in the low excitation transitions of hc@xmath0n .
this is consistent with the fact that both hnc(10 ) and n@xmath4h@xmath126(10 ) , generally considered dense quiescent gas tracers , are found to be very bright towards ic 342 s nucleus @xcite .
hc@xmath0n(54 ) appears to be an excellent extragalactic probe of the dense , quiescent molecular gas component not yet involved in the current starburst . to quantitatively compare star formation with dense gas , a star formation rate
is derived from the 45 ghz continuum flux ( spectral indices measured between 2 cm and 7 mm demonstrate that the vast majority of flux toward gmcs b and c is thermal bremsstrahlung ) . the star formation rate
is then compared with @xmath91 ( table [ tlvg2 ] ) to estimate a dense gas star formation efficiency , sfe@xmath127 .
dense gas depletion timescales , @xmath128= 1/sfe@xmath119 , are also computed . in table
[ tlvg2 ] the ratio of the observed lyman continuum ionization rate ( e.g. * ? ? ?
* ; * ? ? ?
* corrected for distance ) to m@xmath119 derived from the lvg analysis is reported . towards the non - starburst gmcs @xmath129 , while towards gmc b this ratio is 10 - 30 times larger . if one adopts the conversion from n@xmath130 to star formation rate of @xmath131 ( e.g. * ? ? ?
* ) , then sfe@xmath119 for the starburst is @xmath132 , or dense gas depletion times of @xmath133 2 - 3 myr !
this highly enhanced sfe@xmath119 is a direct consequence of the faint hc@xmath0n emission here . even towards the non - starburst clouds sfe@xmath119 are @xmath134 yr@xmath15 .
these efficiencies are sufficiently short that they imply dense gas consumption timescales that are non - negligible fractions of the expected gmc lifetimes .
the dense sfe is rather high across the nucleus , but the extreme value towards gmc b is remarkable .
@xcite argued that intense star formation is suppressed along the spiral arms being triggered when the inflowing molecular gas collides with the inner ring molecular gas .
therefore it is reasonable that in a relative sense sfe@xmath119 is lower away from the central ring .
however this leaves unexplained why gmc b s sfe@xmath119 is so much larger than gmc c , though its positions at the arm / inner ring intersection is the same .
we suggest this is a sign of the evolution of star formation across the nucleus that is impacted by radiative and mechanical feed - back _ from within _ the molecular cloud .
a possible cause of the different sfe@xmath119 between gmc c and b is that we are observing the clouds at high enough spatial resolution to begin to identify the changing internal structure of the clouds in the presence of the starburst . over the lifetime of a cloud sfes
vary . in the earliest stages of a star formation episode sfes will appear low because elevated star formation rates have yet to convert the bulk of the molecular material to stars . towards the final stages of a gmcs evolution
instantaneous sfes appear to increase dramatically as the cloud clumps are consumed , destroyed or dispersed .
so observed instantaneous sfes are expected to vary widely throughout the lifetime of an individual gmc and relative to lifetime averaged sfes typically considered in extragalactic studies .
if starburst b is a few myrs more evolved than the other gmcs , especially the dynamical similar c , then we may be witnessing the consumption , dispersal or destruction phase of the remaining dense clumps in the presence of the expanding hii region .
the magnitude of the dense gas consumption times for b are indeed shorter than the lifetime of the gmc . several lines of evidence suggest that the ( weaker ) star formation towards c2 may be at a somewhat earlier phase .
these include less extended hii regions ( e.g. * ? ? ?
* ) , bright hot core - like species
ch@xmath0oh @xcite , nh@xmath0(6,6 ) @xcite and co(65 ) ( e.g. * ? ? ?
* ) , and more mm dust continuum emission ( e.g. * ? ? ?
* ) . in this context
it is interesting to compare the thermal pressure of the starburst hii region with that of the dense clumps along the same line of sight . assuming a strmgren sphere of r@xmath133 pc , n@xmath135 s@xmath15 and t@xmath136 = 8000 k , parameters determined for the main starburst hii @xcite , the thermal pressure of such an hii region would be @xmath137t@xmath136 = @xmath138 @xmath6 k. this value is equal given the uncertainties to @xmath139t@xmath140 @xmath6 k for the dense component of gmc b , hinting at pressure balance between the dense clumps and the hii region .
in addition , the filling factor of the hii region is larger than implied by the hc@xmath0n lvg analysis .
so it is possible that hc@xmath0n emission towards gmc b comes from dense clumps embedded within the hii region .
the above analysis suggests the following physical picture for the faint hc@xmath0n emission towards the starburst .
the starburst towards b is more evolved .
the hii region at gmc b has had time to expand , destroying or dispersing the dense gas , which is now in the form of smaller clumps and/or more diffuse gas .
clumps that remain there must have high pressure to survive . near the younger star forming cloud gmc c ( particularly toward gmc c2 ) , the hii regions may just be developing , and have not had time to disperse the clouds .
the dense clumps here would be more abundant and still present a hot core - like chemistry .
the clouds further from the central are on average less dense and at the current epoch remain largely quiescent , except possibly d , where the second high excitation component could be associated with the presence of shocks @xcite .
we have imaged the hc@xmath9n j=54 line in the nucleus of ic 342 with the vla and the hc@xmath0n(1615 ) line with the pdbi at @xmath141 resolution .
these are the first maps of these transitions in an external galaxy .
we have detected emission extended along the nuclear mini - spiral " in ( 54 ) and more concentrated emission in ( 16 - 15 ) , with relative abundance of @xmath142 .
_ hc@xmath0n emission is not tightly correlated with star formation strength .
_ dense gas _ excitation _ however , follows star formation more closely .
gmc b , which is weak in all the hc@xmath0n lines , is relatively stronger in the higher j lines .
lvg modeling indicates that the hc@xmath9n - emitting gas has densities of 10@xmath143 . in ic 342 ,
physical conditions of the densest component are fairly constant away from the immediate environment of the starburst , though beyond the central ring densities begin to fall .
comparison with the c@xmath1o observations of @xcite reveal excitation temperatures similar to c@xmath1o values indicating either that the molecular gas is dense and cool ( t@xmath144 k ) or that there are multiple gas components where the densities and kinetic temperatures of each component conspire to give similar overall excitation .
the strong over - prediction of the amount of gas mass present if densities are large and temperatures cool , favors a multi - component ism with at least two components beside the diffuse one seen in @xmath113co(21 ) . the actual ism is likely a continuum of cloud densities with different densities dominating different tracers .
hc@xmath0n also differs in morphology from hcn(10 ) , with hcn(10 ) being much brighter towards the starburst .
this is further evidence that there are multiple dense gas components . of particular interest ,
the starburst site ( gmc b ) exhibits the largest difference in intensity between hc@xmath0n ( both transitions ) and hcn(10 ) .
the faintness of the hc@xmath0n here suggests that the brightness of hcn(10 ) is not due solely to large quantities of dense gas . a comparison between the gmcs with the largest star formation rates and similar dynamical environments , b and c , hint at an explanation .
while gmc b shows higher excitation , the low brightness of this cloud indicates that it is composed of a relatively small amount of warm , dense clumps .
the smaller amount of dense gas at the site of the strongest young star formation indicate high star formation efficiencies in the dense gas . towards gmc c , hc@xmath0n , ch@xmath0oh and nh@xmath0
are more intense and the fraction of millimeter continuum from dust is higher .
this indicates that gmc c is in an early , less evolved ( hot core - like ) state .
the extreme dense gas star formation efficiency observed towards gmc b reflects the fact the main burst is in a more evolved state .
the dense clumps towards the starburst are being dispersed or destroyed in the presence of the hii region .
the little dense gas remaining appears to be in pressure equilibrium with the hii region .
the larger opacity of hcn(10 ) relative to hc@xmath0n elevates its brightness temperature in this warm gas and lowers its critical density permitting it to remain excited in the somewhat lower density component .
we conclude that evla observations of hc@xmath0n(54 ) can be a powerful probe of dense , quiescent molecular gas in galaxies , and when combined with high resolution imaging of the higher j transitions of hc@xmath0n with current and upcoming mm interferometers ( like alma ) provide tight constraints on dense molecular gas properties in stronger or more widespread starbursts , where changes like those localized to gmc b are expected to permeate much of the ism .
we thank chris carilli and miller goss for assistance with the vla observations .
we also thank philippe salome and nemesio rodriguez for their help with the pdbi observations and data reduction .
we thank the referee for several suggestions that improved the presentation . d.s.m .
acknowledges support from the national radio astronomy observatory which is operated by associated universities , inc .
, under cooperative agreement with the national science foundation and nsf grant ast-1009620 .
this work is also supported by nsf grants ast-0307950 and ast-0506469 to j.l.t .
aalto , s. , polatidis , a. g. , httemeister , s. , & curran , s. j. 2002 , , 381 , 783 aladro , r. , martn - pintado , j. , martn , s. , mauersberger , r. , & bayet , e. 2011 , , 525 , a89 becklin , e. e. , gatley , i. , mathews , k. , neugebauer , g. , sellgren , k. , werner , m. k. , & wynn - williams , c. g.1980 , , 236 , 441 brandl , b. r. , et al .
2006 , , 653 , 1129 bussmann , r. s. , et al .
2008 , , 681 , l73 costagliola , f. & aalto , s. , 2010 , , 515 , 71 dahmen , g. , huttemeister , s. , wilson , t. l. , & mauersberger , r. 1998 , , 331 , 959 de vicente , p. , martn - pintado , j. , neri , r. , & colom , p. 2000
, , 361 , 1058 downes , d. , radford , s. j. e. , giulloteau , s. , guelin , m. , greve , a. , & morris , d. 1992 , , 262 , 424 fuente , a. , martin - pintado , j. , cernicharo , j. , & bachiller , r. 1993 , , 276 , 473 gao , y. & solomon , p. m. 2004 , , 606 , 271 garca - burillo , s. , martn - pintado , j. , fuente , a. & neri , r. 2000 , , 355 , 499 garca - burillo , s. , martn - pintado , j. , fuente , a. & neri , r. 2001 , , 563 , l27 garca - burillo , s. , martn - pintado , j. , fuente , a. , usero , a. & neri , r. 2002 , , 575 , l55 graci - carpio , j. , garca - burillo , s. , planesas , p. , fuente , a. , & usero , a. 2008 , , 479 , 703 green , s. & chapman , s. 1978 , , 37 , 169 harris , a. i. , hills , r. e. , stutzki , j. , graf , u. u. , russell , a. g. & genzel , r. , 1991 , , 382 , l75 harrison , a. , henkel , c. , & russell , a. 1999 , , 303 , 157 ho , p. t. p. , martin , r. n. , & ruf , k. 1982 , , 113 , 155 hunter , s. d. et al .
1997 , , 481 , 205 ishizuki , s. , kawabe , r. , ishiguro , m. , okumura , s. k. , morita , k .- i . , chikada , y. , & kasuga , t. 1990 , , 344 , 224 kennicutt , r. c. , jr . 1998 , , 36 , 189 knudsen , k. k. , walter , f. , weiss , a. , bolatto , a. , riechers , d. a. , & menten , k. 2007 , , 666 , 156 krips , m. , neri , r. , garca - burillo , s. , martn , s. , combes , f. , graci - carpio , j. , & eckart , a. 2008 , , 677 , 262 karachentsev , i. d. 2005 , , 129 , 178 lafferty , w. j. & lovas , f. j. 1978 , j. phys .
chem . ref .
data , 7 , 441 larson , r. b. 1981 , , 194 , 809 lo , k. y. et al .
1984 , , 282 , l59 maloney , p. , & black , j. h. 1988 , , 325 , 389 mauersberger , r. , henkel , c. & sage , l. j. 1990 , , 236 , 63 meier , d. s. & turner , j. l. 2001 , , 551 , 687 meier , d. s. , & turner , j. l. 2004 , , 127 , 2069 meier , d. s. & turner , j. l. 2005 , , 618 , 259 meier , d. s. , turner , j. l. & hurt , r. l. 2000 , , 531 , 200 meier , d. s. , turner , j. l. , & hurt , r. l. 2008 , , 675 , 281 montero - castao , m. , herrnstein , r. m. , & ho , p. t. p. 2006
, , 646 , 919 morris , m. , turner , b. e. , palmer , p. & zuckerman , b. 1976 , , 205 , 82 myers , p. c. , & benson , p. j. 1983 , , 266 , 309 narayanan , d. , cox , t. j. , shirley , y. , dav , r. , hernquist , l. , & walker , c. k. 2008 , , 684 , 996 papadopoulos , p. p.
2007 , , 656 , 792 rickard , l. j , & harvey , p. m. 1984 , , 89 , 1520 saha , a. , claver , j. & hoessel , j. g. 2002 , , 124 , 839 sakamoto , s. 1996 , , 462 , 215 schinnerer , e. , bker , t. , & meier , d. s. 2003 , , 591 , l115 schinnerer , e. , bker , t. , meier , d. s. , & calzetti , d. 2008 , , 684 , l21 schulz , a. , gsten , r. , kster , b & krause , d. 2001 , , 371 , 25 scoville , n. z. , & sanders , d. b. 1987 , in interstellar processes , ed .
d. j. hollenbach and h. a. thronson , jr . , ( dordrecht : reidel ) , 21 scoville , n. z. , yun , m. s. , & bryant , p. m. 1997 , , 484 , 702 smith , p. a. , brand , p. w. j. l. , mountain , c. m. , puxley , p. j. , & nakai , n. 1991 , , 252 , 6 solomon , p. m. , rivolo , a. r. , barrett , j. , & yahil , a. 1987 , , 319 , 730 stark , a. a. , bally , j. , wilson , r. w. , & pound , m. w. 1989 , in the center of the galaxy , ed . m. morris , ( dordrecht : kluwer ) , 129 strong , a. w. et al . , 1988 , , 207 , 1 tsai , c .- w . ,
turner , j. l. , beck , s. c. , crosthwaite , l. p. , ho , p. t. p. & meier , d. s. 2006 , , in press turner , j. l. , & ho , p. t. p. 1983 , , 268 , l79 turner , j. l. , hurt , r. l. 1992 , , 384 , 72 turner , j. l. , hurt , r. l. , & hudson , d. y. 1993 , , 413 , l19 usero , a. , garca - burillo , s. , fuente , a. , martn - pintado , j. & rodrguez - fernndez , n. j. 2004 , , 419 , 897 usero , a. , garca - burillo , s. , martn - pintado , j. , fuente , a. , & neri , r. 2006 , , 448 , 457 vanden bout , p. a. , loren , r. b. , snell , r. l. & wootten , a. 1983 , , 271 , 161 wall , w. f. 2007 , , 379 , 674 wu , j. , evans , n. j. , ii , gao , y. , solomon , p. m. , shirley , y. l. , & vanden bout , p. a. 2005 , , 635 , l173 zinchenko , i. , pirogov , l. , & toriseva , m. 1998 , , 133 , 337 lcccccccc a & [email protected] & [email protected] & 22@xmath1452 & 29@xmath1455 & [email protected] & [email protected] & 16@xmath1453 & 44@xmath14510 + b & @xmath125.3 & @xmath121.0 & & & [email protected] & [email protected] & 19@xmath1453 & 28@xmath14510 + c1 & [email protected] & [email protected] & 48@xmath1451 & 39@xmath1453 & [email protected] & [email protected] & 47@xmath1451 & 43@xmath1454 + c2 & [email protected] & [email protected] & 50@xmath1451 & 35@xmath1454 & [email protected] & [email protected] & 45@xmath1451 & 47@xmath1454 + c3 & [email protected] & [email protected] & 49@xmath1451 & 31@xmath1455 & [email protected] & [email protected] & 45@xmath1451 & 47@xmath1454 + d & [email protected] & [email protected] & 51@xmath1456 & 58@xmath14520 & @xmath121.0 & @xmath122.5 & & + d & [email protected] & [email protected] & 52@xmath1452 & 40@xmath1457 & @xmath1461.0 & [email protected] & 57@xmath1458 & 44@xmath14526 + e & [email protected] & 4.4@xmath1451 & 12@xmath1452 & 28@xmath1457 & @xmath121.0 & @xmath122.5 & & + [ tinti ] lcccccccc a & [email protected] & [email protected] & [email protected] & [email protected] & 13@xmath1454 & 4(22 ) & 1(-9 ) & 4.4(5 ) + b & 2.5@xmath1451 & 37@xmath14519 & @xmath530.52 & @xmath5318 & 19@xmath1458 & 4(22 ) & @xmath78(-10 ) & 7.6(5 ) + c1 & [email protected] & [email protected] & [email protected] & [email protected] & 8@xmath1453 & 7(22 ) & 3(-9 ) & 6.8(5 ) + c2 & [email protected] & [email protected] & [email protected] & [email protected] & 8@xmath1453 & 1(23 ) & 2(-9 ) & 1.0(6 ) + c3 & [email protected] & [email protected] & [email protected] & [email protected] & 8@xmath1453 & 5(22 ) & 1(-9 ) & 4.9(5 ) + d & @xmath1460.15 & @xmath135 & @xmath120.17 & @xmath1213 & 6@xmath1454 & 2(22 ) & 3(-9 ) & 6.4(5 ) + d & @xmath130.24 & @xmath136 & [email protected] & [email protected] & @xmath136 & 6(22 ) & 1(-9 ) & 5.8(5 ) + e & [email protected] & [email protected] & @xmath120.057 & @xmath1210 & 7@xmath1453 & 6(22 ) & 1(-9 ) & 1.1(6 ) + [ trat ] lccccccc a & 10 & 5.15 & 0.091 & 38 & 4.7(6 ) & 11 & 6.16 + & 30 & 4.63 & 0.051 & 6.9 & 6.0(5 ) & 1.4 & 6.11 + & 50 & 4.39 & 0.067 & 3.1 & 5.2(5 ) & 1.2 & 6.09 + b & 10 & 5.75 & 0.015 & 75 & 1.3(6 ) & 1.7 & 6.75 + & 30 & 5.22 & 4.7(-3 ) & 14 & 6.5(4 ) & 0.084 & 6.70 + & 50 & 4.87 & 6.0(-3 ) & 5.5 & 4.2(4 ) & 0.055 & 6.57 + & 70 & 4.81 & 5.4(-3 ) & 3.6 & 3.1(4 ) & 0.041 & 6.66 + c1 & 10 & 5.23 & 0.13 & 41 & 9.7(6 ) & 14 & 6.23 + & 30 & 4.69 & 0.066 & 7.4 & 1.0(6 ) & 1.5 & 6.17 + & 50 & 4.50 & 0.075 & 3.6 & 8.0(5 ) & 1.2 & 6.19 + c2 & 10 & 5.36 & 0.12 & 48 & 1.3(7 ) & 13 & 6.36 + & 30 & 4.80 & 0.056 & 8.4 & 1.0(6 ) & 1.0 & 6.28 + & 50 & 4.64 & 0.057 & 4.2 & 7.3(5 ) & 0.73 & 6.34 + c3 & 10 & 5.44 & 0.076 & 53 & 7.1(6 ) & 14 & 6.44 + & 30 & 4.86 & 0.033 & 9.0 & 5.3(5 ) & 1.1 & 6.34 + & 50 & 4.70 & 0.034 & 4.5 & 3.8(5 ) & 0.0.76 & 6.40 + d & 10 & @xmath124.49 & @xmath70.040 & & & & @xmath125.49 + & 30 & @xmath124.10 & @xmath70.040 & & & & @xmath125.58 + & 50 & @xmath123.88 & @xmath70.060 & & & & @xmath125.58 + d & 10 & & & & & & + & 30 & & & & & & + & 50 & & & & & & + e & 10 & 4.90 & 0.12 & 28 & 4.0(6 ) & 3.6 & 5.90 + & 30 & 4.36 & 0.096 & 5.1 & 8.3(5 ) & 0.75 & 5.84 + & 50 & & & & & & + [ tlvg1 ] lccccccc a & 30 & 2.6 & 0.015 & 3.01 & 1.1 & 0.62 & 4.0(45 ) + & 50 & 2.4 & 0.017 & 2.82 & 0.55 & 0.63 & 4.0(45 ) + b & 30 & 2.8 & 3.2(-4 ) & 2.83 & 0.87 & 0.078 & 5.2(46 ) + & 50 & 2.5 & 4.6(-4 ) & 2.54 & 0.37 & 0.097 & 4.2(46 ) + & 70 & 2.4 & 4.0(-4 ) & 2.44 & 0.24 & 0.092 & 4.4(46 ) + c1 & 30 & 2.4 & 0.017 & 3.03 & 1.1 & 0.77 & 1.9(45 ) + & 50 & 2.1 & 0.021 & 2.90 & 0.56 & 0.84 & 1.8(45 ) + c2 & 30 & 2.4 & 0.021 & 3.03 & 1.1 & 0.77 & 2.1(45 ) + & 50 & 2.1 & 0.013 & 2.87 & 0.54 & 0.83 & 1.9(45 ) + c3 & 30 & 2.3 & 5.0(-3 ) & 2.80 & 0.84 & 0.69 & 4.1(45 ) + & 50 & 2.1 & 6.3(-3 ) & 2.64 & 0.42 & 0.72 & 4.0(45 ) + d & 30 & @xmath122.5 & & & & & + & 50 & @xmath122.3 & & & & & + d & 30 & & & & & & + & 50 & & & & & & + e & 30 & 2.3 & 0.030 & 2.94 & 0.99 & 0.78 & 1.6(45 ) + & 50 & 2.0 & & & & & + [ tlvg2 ] | we present the first images of the j=54 and j=1615 lines of the dense gas tracer , cyanoacetylene , hc@xmath0n , in an external galaxy .
the central 200 pc of the nearby star - forming spiral galaxy , ic 342 , was mapped using the vla and the plateau de bure interferometer .
hc@xmath0n(54 ) line emission is found across the nuclear mini - spiral , but is very weak towards the starburst site , the location of the strongest mid - ir and radio emission .
the j=1615 and 109 lines are also faint near the large hii region complex , but are brighter relative to the 54 line , consistent with higher excitation .
the brightest hc@xmath0n emission is located in the northern arm of the nuclear minispiral , 100 pc away from the radio / ir source to the southwest of the nucleus .
this location appears less affected by ultraviolet radiation , and may represent a more embedded , earlier stage of star formation .
excitation temperatures are consistent with those determined from c@xmath1o ; the gas is dense , @xmath2 and cool , t@xmath3 k. so as to not violate limits on the total h@xmath4 mass determined from c@xmath1o , at least two dense components are required to model ic 342 s giant molecular clouds .
these observations suggest that hc@xmath0n(54 ) is an excellent probe of the dense , quiescent gas in galaxies .
the high excitation combined with faint emission towards the dense molecular gas at the starburst indicates that it currently lacks large masses of very dense gas .
we propose a scenario where the starburst is being caught in the act of dispersing or destroying its dense gas in the presence of the large hii region .
this explains the high star formation efficiency seen in the dense component .
the little remaining dense gas appears to be in pressure equilibrium with the starburst hii region . |
multimedia images are widely used in internet communications , so the need for securely transmitting these images over networks has become an essential part of the field of data security . image cryptography plays a vital role in securing confidential images . the purpose of image cryptography is to hide the content of the images by encrypting them so as to make the images unrecognizable to the intruders .
one part of this paper deals with a new image cryptosystem using 2d cellular automata .
+ in general , there are two different methods to protect an image ; they are ( i ) image shuffling and ( ii ) image encryption .
pixels positions are rearranged in image shuffling whereas in image encryption , pixel values and positions are changed . in both the cases it is very essential to check the security of the method .
that means the method should be invulnerable to all attacks .
poorly protected images will always provide information about the original image in statistical analysis . if the encrypted image is indistinguishable from the random image , statistical analyses will not have advantage to break . so testing the randomness in the pixels of encrypted image is the state of the art . in the literature
already there are different tests for checking randomness for 1d data@xcite .
a number of parametric tests are designed for pixel randomness in shuffled and encrypted images@xcite .
a non - parametric test is developed in this paper for checking randomness in the image pixels , which is the first of its kind . + john von neumann proposed a new emerging concept called cellular automata ( ca)@xcite .
ca is a discrete model consisting of regular grid of cells , each in one of the finite number of states .
according to some fixed rule , the state of each cell will be changed in terms of the state of the current cell and the states of the cells in its neighborhood .
like 1d , higher dimension ca also can be defined@xcite .
it is already proved that some of the rules of ca will be able to generate complex random patterns@xcite . in the last two decades 1d and 2d
ca s are used in cryptography .
lot of research is going on for ca based image cryptography @xcite .
the paper is organized as follows : section 2 describes the basics of 2d cellular automata and the 2d ca concepts used for our encryption scheme .
section 3 presents our new non - parametric test for pixel randomness .
section 4 describes our 2d ca based encryption scheme .
section 5 presents the simulation results and performance evaluations .
section 6 gives the concluding remarks .
the extensional behavior of 1d cellular automata is also able to produce complicated patterns in two - dimension .
this extension is significant since this is compared with pattern formation in physical systems .
2d ca is a regular 2d lattice of cells .
each cell has n possible values and is updated in each discrete time steps according to a rule f that depends on the value of sites in some neighbor around it .
there are different types of lattices and neighborhood structures in a 2d ca .
figure 1 shows the two familiar neighborhood structures named as von - neumann neighborhood and moore neighborhood .
the value of @xmath0 of a cell at @xmath1th position in @xmath2th time in a 2d ca with a rule f , that depends only on the cells according to von - neumann neighborhood is evolved from @xmath3 .
+ ( * a * ) ( -1.9,-1.9 ) grid ( 1.4,1.4 ) ; ( -0.5,-0.5 ) rectangle ( 0,0 ) ; ( -0.5,0 ) rectangle ( 0,0.5 ) ; ( 0,-0.5 ) rectangle ( 0.5,0 ) ; ( -0.5,-1 ) rectangle ( 0,-0.5 ) ; ( -1,-0.5 ) rectangle ( -0.5,0 ) ; ( * b * ) ( 2.1,-1.9 ) grid ( 5.4,1.4 ) ; ( 3.5,-0.5 ) rectangle ( 4,0 ) ; ( 3,0 ) rectangle ( 4.5,0.5 ) ; ( 3,-1 ) rectangle ( 3.5,0 ) ; ( 3.5,-1 ) rectangle ( 4.5,-0.5 ) ; ( 4,-0.5 ) rectangle ( 4.5,0 ) ; + 1d and 2d ca are studied differently with the help of polynomial algebra and matrix algebra by researchers .
p.p.choudary et.al.already have studied and designed a new characteristic of 2d ca@xcite .
these concepts are used for our proposed encryption method .
2d nine neighborhood moore s neighborhood is used here for evolutions .
different rules are defined in terms of dependencies from the following rule convention .
+ [ cols="^,^,^,<",options="header " , ] * table 2 : * scores of test images .
we have introduced a new test using non - parametric method in statistics for measuring the randomness in the image pixels .
it scores the images , depending upon how far their pixels are i.i.d .
the newly designed 2d ca based encryption scheme first analysed by standard evaluation methods and then the encrypted images are used to validate our newly proposed non - parametric test .
simulation results clearly show that this test can be used for evaluating the quality of any image shuffling and image encryption method.the complexity of the method is @xmath4 , so it can be efficiently implemented . in the proposed non - parametric test we have only used the basic run test to measure the randomness .
balasuyambu jeyaram , rama raghavan , krishna shankara narayanan , new ca based key generation for a robust rgb color image encryption scheme , international journal of computer applications(0975 8887 ) , volume 80-no.7 , october 2013 , pp.45 - 50 .
balasuyambu jeyaram , rama raghavan , new ca based image encryption - scaling scheme using wavelet transform , journal of systemics , cybernatics and informatics , volume 12- number 3- year 2014 , issn : 1690 - 4524 , pp.66 - 71 .
p. p. choudhury , birendra kumar nayak , sudhakar sahoo , sunil pankaj rath , theory and applications of two - dimensional , null - boundary , nine - neighborhood , cellular automata linear rules .
corr abs/0804.2346 2008 .
p. chattopadhyay , p. p. choudhury , characterisation of a particular hybrid transformation of two - dimensional cellular automata , computers and mathematics with applications , elsevier publication , 38 , 1999 , pp.207 - 216 .
k. dihidar , p. p.
choudhury , matrix algebraic formulae concerning some special rules of two - dimensional cellular automata , international journal on information sciences , elsevier publication , volume 165 , 2004 , pp.91 - 101 . j. d. gibbons and s. chakraborti , nonparametric statistical inference , new york : marcel dekker , 1992 .
li , x. , `` a new measure of image scrambling degree based on grey level difference and information entropy . '' , international conference on computational intelligence and security , ieee,1 , 2008,pp.350 - 354
. n. h. packard and s. wolfram , two - dimensional cellular automata , journal of statistical physics , 38 ( 5/6 ) , 1985 , pp.901 - 946 . w. pries , a thanailakis and h. c. card , group properties of cellular automata and vlsi applications , ieee trans on computers c-35 , december 1986 , pp.1013 - 1024 .
r , bala suyambu . j , a. arokiaraj , s. saravanan , a study of des algorithm with cellular automata , international journal of innovative management , information and production , volume 3 , number 1 , march 2012 , isme international 2011 , issn 2185 - 5439 , pp.10 - 16 .
rukhin , a. , soto , j. , nechvatal , j. , smid , m. , barker , e. , leigh , s. , leveson , m. , banks , d. , heckert , a. , dray , j. , and vo , s. , `` a statistical test suit for random and pseudorandom number generators for cryptographic applications , '' nist special publication , 800 - 22 , 2010 .
neumann , the theory of self - reproducing automata , ( edited by a. w. burks ) univ . of illinois press urbana , 1996 .
s. wegenkittl , `` entropy estimators and serial tests for ergotic chains . '' , ieee trans .
theory , vol .
2001 , pp.2480 - 2489 .
s.wolfram , statistical mechanics of cellular automata , rev mod phys .
55 , july 1983 , pp.601 - 644 .
wu , y. , zhou , y. , saveriades , g.,agaian , s. , noonan , j. , and natarajan , p. , `` local shanon entropy measure with statistical tests for image randomness '' , elsevier publication , information sciences 2012 .
yue wu , sos agaian and joseph p. noonan , `` a novel method of testing image randomness with applications to image shuffling and encryption '' proc .
spie 8755 , mobile multimedia / image processing , security and applications , 2013 . | in this paper we have proposed a new test for pixel randomness using non - parametric method in statistics . in order to validate this new non - parametric test
we have designed an encryption scheme based on 2d cellular automata .
the strength of the designed encryption scheme is first assessed by standard methods for security analysis and the pixel randomness is then determined by the newly proposed non - parametric method .
* keywords : * 2d cellular automata , image encryption , pixel randomness , non - parametric test . |
since globally defined lipschitz mappings have good rademacher type differentiability properties , it is a matter of general interest to provide extension theorems for partially defined lipschitz mappings . in this context , the following concept is useful : a pair of metric spaces @xmath4 is said to have the _
lipschitz extension property _
if there exists a constant @xmath5 such that any lipschitz mapping @xmath6 , @xmath7 , has a lipschitz extension @xmath8 with the property that @xmath9 . here , we used the notation @xmath10 and similarly for @xmath11 .
there is a long history of results providing examples of metric pairs @xmath4 with or without the lipschitz extension property .
the first such result is due to mcshane @xcite who proved that for an arbitrary metric space @xmath12 , the pair @xmath13 has the lipschitz extension property , even with constant @xmath14 .
this result was followed by the theorem of kirszbraun @xcite who proved the lipschitz extension property , also with constant @xmath14 , for pairs of euclidean spaces @xmath15 .
kirszbraun s theorem was generalized by valentine @xcite to pairs of general hilbert spaces @xmath16 .
the lipschitz extension problem becomes more challenging if the target space @xmath17 has no linear structure .
an generalization of the kirszbraun
valentine theorem to metric spaces with a curvature bound in the sense of alexandrov was established by lang and schroeder @xcite , and the case of an arbitrary metric space @xmath12 and a hadamard manifold @xmath17 was discussed in @xcite . in the paper of lang and schlichenmaier @xcite , a general sufficient condition for the lipschitz extension property of a metric pair @xmath4 is given .
the lipschitz extension property for a pair of general carnot groups @xmath18 with their canonical sub - riemannian carnot - carathodory metrics is studied for particular pairs of carnot groups in @xcite , @xcite , @xcite .
the situation is fully understood for the pair @xmath19 , where @xmath20 is the euclidean space of dimension @xmath2 and @xmath21 is the heisenberg group of topological dimension @xmath22 . in this case , it is known that the pair @xmath19 has the lipschitz extension property if and only if @xmath23 . the `` only if '' part of this equivalence is due to balogh and fssler @xcite , whereas the `` if '' part is due to wenger and young @xcite . in this note , we shall study the lipschitz extension property for pairs of heisenberg groups @xmath24 , a question that was posed in @xcite .
let us observe first that since @xmath0 contains isometric copies of @xmath25 , it follows from @xcite that if @xmath24 has the lipschitz extension property , then we must necessarily have that @xmath23 .
the reason of this restriction on dimensions is as follows : it is a well - known fact of sub - riemannian geometry that smooth lipschitz mappings between sub - riemannian spaces must satisfy the _ contact condition_. this means that the horizontal distribution defining the sub - riemannian metric must be preserved by the tangent map .
this property gives an obstruction for the existence of smooth lipschitz extensions .
this principle can be transferred even for possibly non - smooth lipschitz extensions , allowing us to prove the main result of this note : [ main ] if @xmath26 , then the pair @xmath27 does not have the lipschitz extension property . since @xmath28 is isometrically
embedded in @xmath0 for @xmath29 it is enough to prove the statement for @xmath30 .
the idea of the proof is to exploit the consequences of the aforementioned contact condition . indeed
, this condition implies that a globally defined smooth contact map @xmath31 is orientation preserving , which gives a major restriction on the nature of globally defined lipschitz mappings .
it allows us to define a map @xmath32 from a subset @xmath33 into @xmath28 that does not have a globally defined lipschitz extension . in order to prove the non - existence of such an extension we use results from mapping degree theory @xcite , @xcite and pansu s result on almost everywhere differentiability @xcite
this argument does not work for the case of the pairs @xmath34 as smooth contact maps @xmath35 are not necessarily orientation preserving .
in this case the lipschitz extension property is still open .
the paper is organized as follows : in section 2 we recall the metric structure of the heisenberg group and discuss regularity results of lipschitz maps between heisenberg groups to be used in the sequel . in section 3
we prove a result on the computation of the mapping degree of maps between heisenberg groups using the pansu derivative .
this result may be of independent interest .
in section 4 we give the proof of theorem [ main ] .
the authors thank jeremy t. tyson , stefan wenger , piotr hajlasz and enrico le donne for many stimulating conversations on the subject of the paper .
we recall background results on the geometry of the heisenberg group and preliminary results on the differentiability properties of lipschitz maps between heisenberg groups .
the model of the heisenberg group @xmath0 used in this paper is described as follows : @xmath36 , endowed with the group operation @xmath37 for @xmath38 . here ,
@xmath39 stands for the usual scalar product of vectors @xmath40 .
it is easy to check that the vector fields @xmath41 are left invariant for @xmath42 .
for @xmath43 the space @xmath44 is the horizontal plane at @xmath45 .
varying @xmath43 gives rise to a vector bundle @xmath46 , the horizontal plane distribution .
let us choose the riemannian metric on @xmath46 which makes @xmath47 orthonormal at each point @xmath43 and denote the length of horizontal vectors @xmath48 in this metric by @xmath49 .
a lipschitz curve @xmath50 \to { { \mathbb h}}^n$ ] is said to be horizontal if @xmath51 for a.e .
@xmath52 $ ] . for @xmath53
we denote by @xmath54 the sub - riemannian carnot - carathodory distance between @xmath45 and @xmath55 defined as @xmath56 where the @xmath57 is taken over all horizontal curves connecting @xmath45 and @xmath55 .
such horizontal curves exist between any pair of points @xmath53 .
this follows from the fact that the horizontal distribution is bracket - generating . for an explicit computation of the geodesics and the exact formulae of the metric we refer to @xcite
( see also @xcite or @xcite ) .
it is easy to check that @xmath58 is invariant under left translation .
moreover , it is compatible with the homogeneous dilations @xmath59 defined by @xmath60 , i.e. @xmath61 for all @xmath62 and @xmath53 . in this paper the metric on @xmath0
is understood to be the above defined carnot - carathodory metric @xmath58 .
however the main result of the paper remains valid for any metric @xmath63 on @xmath0 that is bi - lipschitz equivalent to @xmath58 .
it is easy to see that a smooth lipschitz map @xmath64 has the property that its tangent map preserves the horizontal bundle @xmath46 . since the horizontal bundle is the kernel of the one - form @xmath65 we obtain that there exists a real valued smooth function @xmath66 such that @xmath67 .
if @xmath68 is merely lipschitz with no additional smoothness assumptions , then @xmath66 is only an @xmath69 function and the equality @xmath67 holds almost everywhere .
this follows from the following differentiability result : a lipschitz map @xmath70 is called @xmath71-differentiable at a point @xmath72 if there exists a group homomorphism @xmath73 such that @xmath74 for all @xmath75 .
here @xmath76 stands for the group operation , @xmath77 is the homogeneous group dilation , and @xmath78 is the inverse to @xmath45 in @xmath79 .
pansu proved in @xcite that any lipschitz map is @xmath71-differentiable almost everywhere , moreover in a point of @xmath71-differentiability the limit is uniform in @xmath55 if @xmath55 is taking its values in a compact set .
let us define the singular set of a lipschitz map @xmath80 as @xmath81 @xmath82 @xmath83 and the regular set as @xmath84 for @xmath85 we denote by @xmath86 the @xmath63-dimensional hausdorff measure with respect to the heisenberg metric .
by pansu s theorem we have @xmath87 , which implies by the lipschitz continuity of @xmath68 that also @xmath88 .
furthermore , we observe that if @xmath89 , the metric jacobian @xmath90 of @xmath68 at @xmath45 vanishes , where @xmath91 and @xmath92 is the ( open ) ball with center @xmath45 and radius @xmath93 in the heisenberg metric .
for any measurable set @xmath94 we have @xmath95 ( see e.g. chapter 6 of @xcite ) , so @xmath96 and @xmath97
in this section we will use the @xmath71-derivative in order to estimate the mapping degree of lipschitz maps between heisenberg groups . for background results on degree theory
we refer to @xcite , @xcite .
the considerations of this section are based on the fact that smooth contact maps @xmath98 have a positive jacobi determinant , and so they preserve orientation . to see this ,
let @xmath99 be the contact form in @xmath28 .
observe that if we denote by @xmath100 the volume form in @xmath28 , then there exists a dimensional constant @xmath101 such that @xmath102 .
if @xmath68 is a smooth contact mapping then @xmath103 for some real valued smooth function @xmath66 .
the relation between the jacobi determinant of @xmath68 and @xmath66 can be computed as follows : @xmath104 which implies that @xmath105 . in particular
, it follows that group homomorphisms @xmath106 are orientation preserving .
we want to extend this observation from the class of smooth lipschitz maps to possibly non - smooth lipschitz mappings .
this is possible by a version of the rademacher differentiability theorem due to pansu @xcite according to which lipschitz maps between carnot groups are differentiable almost everywhere ; and their tangent map satisfies the contact condition in the points of differentiability .
note that differentiability is understood here in terms of the underlying group operation and compatible homogeneous dilations as indicated in the previous section .
the classical rademacher differentiability theorem that is formulated in terms of vector addition and scalar multiplication as in the euclidean space fails for lipschitz maps between carnot groups .
an example showing this fact is due to magnani @xcite .
nevertheless , pansu s differentiability theorem can be used to estimate the mapping degree for maps between carnot groups .
the main result of this section is the following statement : [ degree ] let @xmath107 be an integer , @xmath98 a lipschitz map , @xmath108 an open set , and @xmath109 . if @xmath110 , then @xmath111 let us consider a ( possibly ) non - smooth lipschitz map @xmath98 as in the statement .
there are two cases to consider : if @xmath112 , then @xmath113 by the definition of the degree ( @xcite , @xcite ) and we are done .
let us therefore assume that @xmath114 $ ] . in this case we shall prove that @xmath115 the proof is based on the fact that for group homomorphisms this property holds true as indicated above .
this fact will be combined with a localization argument using linearization by the @xmath71-differential of @xmath68 at regular points .
the first step is to show that if @xmath114 $ ] , the pre - image @xmath116 is a finite set . to prove this , assume by contradiction that there exists an infinite sequence @xmath117 of distinct points in @xmath118 such that @xmath119 for each @xmath120 . by passing to a subsequence
we can assume that @xmath121 for some @xmath122 .
by continuity of @xmath68 we obtain that @xmath123 .
since @xmath124 , this implies that @xmath125 .
we shall obtain the desired contradiction by showing that @xmath126 is non - injective . to see this ,
let us write @xmath127 for some @xmath128 and @xmath129 , @xmath130 . by passing to a further subsequence
if necessary we can assume that @xmath131 for some @xmath132 . since @xmath133 , we have @xmath134 but this last limit equals @xmath135 by the lipschitz continuity of @xmath68 and the fact that @xmath131 .
so @xmath136 , in contradiction to @xmath137 .
by the first step we can assume that @xmath138 by general properties of the mapping degree ( @xcite , @xcite ) we have that @xmath139 where @xmath140 are arbitrary small , non - intersecting open sets containing @xmath141 for @xmath142 . in the following we shall prove that @xmath143 for every @xmath144 .
the idea is to construct a homotopy from @xmath68 to its linearization @xmath145 in @xmath146 with the property that the image of @xmath147 under the homotopy does not meet the point @xmath45 .
let us fix an index @xmath148 . to simplify the notation ,
we can assume without loss of generality that @xmath149 , @xmath150 for some small @xmath151 ( to be chosen later ) , and @xmath152 , where @xmath153 is an injective homogeneous group homomorphism ( i.e. commuting with dilations ) .
the @xmath71-differentiability of @xmath68 at @xmath135 now reads as @xmath154 and the convergence is uniform in @xmath132 .
equivalently , for every @xmath155 there exists an @xmath156 such that @xmath157 for all @xmath158 and @xmath159 .
furthermore , by the injectivity and homogeneity of @xmath160 there exists an @xmath161 such that @xmath162 for all @xmath132 and @xmath62 .
let us choose @xmath163 and @xmath164 for some @xmath165 .
consider the homotopy @xmath166 \times \overline{b(0,r ) } \to { { \mathbb h}}^{2k+1}$ ] given by @xmath167.\ ] ] clearly @xmath168 and @xmath169 .
for all @xmath170 $ ] and @xmath171 we have @xmath172 where @xmath173 by and @xmath174 by , so that @xmath175 by the homotopy invariance property of the mapping degree ( @xcite , @xcite ) we obtain that @xmath176 finishing the proof .
observe that the statement of proposition [ degree ] no longer holds for maps @xmath177 . indeed
, the mapping @xmath177 given by @xmath178 is a contact map since @xmath179 , and it is orientation reversing : @xmath180 . more generally , it is easy to see that every integer is the degree of some contact map between domains in @xmath181 .
in this section we define a lipschitz mapping on a subset @xmath182 of @xmath79 to itself which does not have a lipschitz extension to @xmath79 in the case when @xmath30 .
the set @xmath182 will be a topological @xmath183-sphere foliated by geodesics connecting the origin @xmath184 to a point @xmath185 on the vertical axis .
for simplicity of notation we choose @xmath186 .
to define @xmath182 we use a parametrization @xmath187 \to { { \mathbb h}}^{n}$ ] , where @xmath188 is the unit sphere in @xmath189 . the map @xmath187 \to { { \mathbb h}}^{n}$ ] will be given by @xmath190 for @xmath191 $ ] .
the hypersurface @xmath192)$ ] is the boundary of a bounded domain @xmath193 and is foliated by the unit speed geodesics @xmath194 connecting @xmath184 and @xmath185 .
for details we refer to the papers of ambrosio - rigot @xcite and monti @xcite .
we claim that to each point @xmath195 there is a unique value @xmath196 such that @xmath197 , whereas @xmath198 and @xmath199 . to verify this ,
. then @xmath201 , and since the function @xmath202 , @xmath203 is strictly increasing and onto , we get a unique value @xmath204 such that @xmath205 . furthermore , the system @xmath206 has a unique solution @xmath207 , proving the claim .
consider now the _
symplectic reflection _ @xmath208 given by @xmath209 .
observe that @xmath210 for the standard symplectic form @xmath211 .
we define the mapping @xmath212 using @xmath213 as follows : if @xmath214 , set @xmath215 .
if @xmath216 , let @xmath217 and put @xmath218 observe that the mapping interchanges the geodesics according to the reflection @xmath219 while it preserves the _ height _
@xmath220 of points on @xmath182 .
[ example ] the mapping @xmath212 defined above is a lipschitz map with respect to the restriction of the sub - riemannian carnot - carathodory metric @xmath58 on @xmath182 in both the source and target space .
we have to find a constant @xmath221 such that @xmath222 let @xmath223 .
we can assume that @xmath224 , so that there are unique preimages @xmath217 and @xmath225 . consider a third point @xmath226 lying on the same geodesic as @xmath227 but having the height of @xmath45 , that is , @xmath228 . by the triangle inequality we have @xmath229 since @xmath55 and @xmath227 are on the same geodesic , the same is true for @xmath230 and @xmath231 .
it follows that @xmath232 hence we obtain @xmath233 we claim that there is a constant @xmath234 such that @xmath235 here we will use in an essential way that @xmath45 and @xmath55 have the same heights . in order to verify the above estimate we shall consider another metric @xmath236 that is bi - lipschitz equivalent to @xmath58 and prove that @xmath237 the estimate follows from and the bi - lipschitz equivalence of @xmath58 and @xmath236 . for two points @xmath238 the kornyi gauge metric @xmath236
is given by the expression @xmath239 where @xmath240 and @xmath241 denote the euclidean norm and scalar product in @xmath242 .
it can be checked that this defines a metric indeed .
this metric is no longer given as a sub - riemannian metric but is bi - lipschitz equivalent to @xmath58 .
we refer to @xcite for details .
to verify equality we shall use the explicit forms of @xmath45 and @xmath55 : @xmath243 where @xmath204 and @xmath244 .
the respective expressions for @xmath245 and @xmath230 are obtained by replacing @xmath246 by @xmath247 as well as @xmath248 by @xmath249 and leaving @xmath250 and @xmath251 unchanged .
now a direct computation gives @xmath252 to calculate @xmath253 , all we need to do is to replace on the right side of @xmath246 by @xmath247 and @xmath248 by @xmath249 . after doing that
, we see that its value will not change .
this gives and hence .
notice that in this computation it was crucial that the two points @xmath45 and @xmath55 have the same heights .
otherwise , the second part of the right side of will be different .
we are now ready to prove our main result about the existence of partially defined lipschitz maps on @xmath28 with no lipschitz extension to the whole heisenberg group @xmath28 .
this follows from the following : by contradiction assume that there is a lipschitz extension @xmath259 such that @xmath258 .
proposition [ degree ] implies that if @xmath30 , then @xmath260 let us choose @xmath261 such that @xmath262 and let @xmath45 be any point in the interior of @xmath263 , for instance , take @xmath264 . since @xmath265 is a reflection on the boundary we have @xmath266 which gives a contradiction , proving the statement .
we will use the relation first to find an explicit formula for the mapping @xmath212 in cartesian coordinates as @xmath271,\ ] ] where @xmath272 \to [ 0,2\pi]$ ] is the inverse function of @xmath273 \to [ 0,4\pi]$ ] , @xmath274 .
observe now that @xmath275 is a smooth function and formula itself in fact gives a continuous extension @xmath276 that is smooth in @xmath263 .
the above formula indicates that @xmath277 preserves horizontal planes . in horizontal planes
the map looks like a reflection followed by a rotation where the rotation angle depends on the height of the horizontal plane . since @xmath278 is independent on the choice of the extension we have @xmath279 .
because the mapping @xmath277 is smooth in @xmath263 we can calculate the degree by an explicit computation as @xmath280 in the case when @xmath30 we obtain that @xmath281 , finishing the proof . in conclusion ,
let us come back to the general problem of determining exactly the pairs of heisenberg groups @xmath24 with the lipschitz extension property .
consider the case @xmath282 . in
@xcite it is shown that the pair @xmath283 does have the lipschitz extension property .
on the other hand , by theorem [ main ] this is not case for @xmath284 .
therefore it only remains to study the extension problem for the pair @xmath285 .
we believe that understanding the lipschitz extension property in this particular situation will shed light also on the general case of pairs @xmath24 .
optimal mass transportation in the heisenberg group , _ j. functional analysis _ 208 ( 2004 ) , 261301 .
rectifiability and lipschitz extensions into the heisenberg group , _ math .
z. _ 263 ( 2009 ) , 673683 .
an introduction to the heisenberg group and the sub - riemannian isoperimetric problem , birkhuser , 2000 .
degree theory in analysis and applications , oxford science publications , 1995 .
ber die zusammenziehende und lipschitzsche transformationen , _ fundam .
_ 22 ( 1934 ) , 77108 .
agata dimension , quasisymmetric embeddings , and lipschitz extensions , _ int .
res . not .
_ 58 ( 2005 ) , 36253655 .
extensions of lipschitz maps into hadamard spaces , _ geom .
_ 10 ( 2000 ) , 15271553 .
kirszbraun s theorem and metric spaces of bounded curvature , _ geom .
_ 7 ( 1997 ) , 535560 . degree theory , cambridge university press , 1978 .
the coarea formula for real - valued lipschitz maps on stratified groups , _ math .
_ 278 ( 2005 ) , 16891705 .
extension of range of functions , _ bull .
_ 40 ( 1934 ) , 837842 . some properties of carnot - carathodory balls in the heisenberg group , _ rend
lincei _ 11 ( 2000 ) , 155167 .
mtriques de carnot - carathodory et quasiisomries des espaces symtriques de rang un , _ ann . math .
_ 129 ( 1989 ) , 160 .
lipschitz non - extension theorems into jet space carnot groups , _ int .
res . not .
_ 18 ( 2010 ) , 36333648 . a lipschitz condition preserving extension for a vector function , _ amer
. j. math .
_ 67 ( 1945 ) , 8393 .
lipschitz extensions into jet space carnot groups , _ math .
_ 17 ( 2010 ) , 11371149 . | let @xmath0 be the heisenberg group of topological dimension @xmath1 . we prove that if @xmath2 is odd , the pair of metric spaces @xmath3 does not have the lipschitz extension property . |
van der waals ( vdw ) forces are non - vanishing forces between neutral , unpolarized objects that arise as a consequence of quantum ground - state fluctuations of the electromagnetic field and the interacting objects . the first theory of vdw forces within the framework of full quantum electrodynamics ( qed ) goes back to the pioneering work of casimir and polder in 1948 @xcite .
using a normal - mode expansion of the quantized electromagnetic field , they derived the vdw force on a ground - state atom in front of a perfectly conducting plate as well as that between two ground - state atoms in free space .
whereas the normal - mode technique requires specification of the geometry of the system at a very early point of the calculation , alternative approaches based on linear response theory allow for obtaining geometry - independent results . employing the dissipation - fluctuation theorem ,
geometry - independent expressions for the vdw potentials of one and two ground - state atoms in terms of the ( classical ) green tensor of the macroscopic maxwell equations and the polarization of the atom can be derived @xcite results that can be confirmed @xcite within the framework of exact macroscopic qed in linear , causal media @xcite .
the abovementioned methods apply to atoms in free space only , because the local electromagnetic field acting on an atom situated in a medium differs from the macroscopic one due to the influence of the local medium environment of the atom . within macroscopic qed ,
one approach to overcome this problem and to account for the local - field correction is the real - cavity model @xcite , where the guest atom is assumed to be surrounded by an small , empty , spherical cavity .
the change of the electric field due to the presence of the cavity is then expressed via the associated green tensor that characterizes the local environment of the guest atom .
the real - cavity model has been applied to the spontaneous decay of an excited atom in a bulk medium @xcite and at the center of a homogeneous sphere @xcite .
the local - field corrected decay rate for the latter case was found to be the uncorrected rate multiplied by a global factor and shifted by a constant term .
it was suggested that the result , reformulated in terms of the associated green tensor , remains valid for arbitrary geometries .
this was proven correct in ref .
@xcite , where it it is shown that the local - field corrected green tensor can under very general assumptions be decomposed into the sum of the uncorrected green tensor modified by a factor and a translationally invariant term .
such a decomposition of the green tensor is not only appropriate to obtain the abovementioned results for the rate of spontaneous decay .
it can also be used to calculate local - field corrected vdw potentials .
a first hint was given in ref .
@xcite , where the local - field corrected vdw interaction of two ground - state atoms placed in adjacent semi - infinite magnetoelectric media is studied .
van der waals interactions play an important role in various fields , where often the interacting objects are embedded in media so that local - field effects may be expected to play a major role .
for example , in colloid science the mutual vdw attractions of colloidal particles suspended in a liquid influence the stability of such suspensions @xcite .
in particular , clustering of particles ( flocculation ) may occur unless the suspension is sufficiently balanced by electrostatic repulsive forces @xcite .
examples of vdw interactions of microobjects embedded in media can also be found in biology @xcite , such as cell cell , cell substratum , cell virus , and cell vesicle interactions @xcite , where the investigations are usually based on the works of lifshitz @xcite and hamaker @xcite .
so , it has been studied whether the theory of vdw interactions can explain the phenomenon of biomolecular organization @xcite .
in particular , electrostatic and vdw forces between a protein crystal and a molecule in solution have been ( theoretically ) compared @xcite , and the non - additivity of vdw interactions between two layers within a multilayer assembly such as lipid
water has been investigated @xcite .
needless to say that vdw interactions represent only one effect among others and that the systems are usually very complex , involving a broad range of issues such as surface charges , simultaneous existence of many different molecules , flexibility of biological macromolecules , fluidity of membranes , and various types of non - covalent forces , e.g. , vdw , electrostatic , solvation , steric , entropic , and structural ones @xcite . in the present paper , single- and many - atom
vdw interactions of ground - state atoms embedded in media are studied within the framework of macroscopic qed , where the local - field correction is included via the real - cavity model .
the paper is organized as follows . in sec .
[ sec_basic ] , basic formulas for the vdw potentials of ground - state atoms are reviewed .
section [ sec_real ] is concerned with the real - cavity model , where a general decomposition of the local - field corrected green tensors entering the single- and two - atom vdw potentials is presented .
the results are combined in sec .
[ sec_local ] with the general formulas for the vdw potential as given in sec .
[ sec_basic ] . closed expressions for the local - field corrected single- and many - atom vdw potentials are presented and briefly discussed . as an application , the vdw potential is calculated for two atoms placed within a magnetoelectric bulk material .
some concluding remarks are given in sec .
the vdw potential of a ( non - magnetic ) atom can be identified with the position - dependent part of the shift of the ( unperturbed ) eigenenergy , which arises from the interaction of the atom with the medium - assisted electromagnetic field in the ground state . in leading - order perturbation theory ,
the vdw potential of a ground - state atom @xmath0 at position @xmath1 in the presence of a linearly responding , dispersing , and absorbing medium , can be given in the form @xmath2\ ] ] ( for a derivation within the framework of macroscopic qed , see , e.g. , ref .
@xcite ) . here
, @xmath3 is the atomic ground - state polarizability tensor in lowest non - vanishing order of perturbation theory ( see , e.g. , ref .
@xcite ) , with @xmath4 @xmath5 @xmath6 @xmath7 @xmath8 and @xmath9 @xmath5 @xmath10 , respectively , being the ( unperturbed ) atomic transition frequencies and the atomic electric - dipole transition matrix elements .
further , the scattering part of the ( classical ) green tensor @xmath11 of the electromagnetic field accounts for scattering at inhomogeneities of the medium .
recall that the green tensor of the entire system can be decomposed according to @xmath12 [ @xmath13 , bulk part ] .
in particular for locally responding , isotropic , magnetoelectric matter , the green tensor satisfies the differential equation @xmath14 { \mbox{\textbf{\textsf{g}}}}(\mathbf{r},\mathbf{r'},\omega)= \delta(\mathbf{r}-\mathbf{r'}){\mbox{\textbf{\textsf{i}}}}\ ] ] ( @xmath15 , unit tensor ) together with the boundary condition @xmath16 the ( conservative ) force @xmath17 associated with the potential ( [ potone ] ) can be obtained according to @xmath18 next , consider two neutral , unpolarized ( non - magnetic ) ground - state atoms @xmath0 and @xmath19 in the presence of a linearly responding medium .
the force acting on atom @xmath0 due to the presence of atom @xmath19 can be derived from the two - atom potential obtained in fourth - order perturbation theory as ( cf .
@xcite ) @xmath20.\end{gathered}\ ] ] note that in contrast to eq .
( [ potone ] ) , the full green tensor @xmath21 comes into play .
the force @xmath22 on the atom @xmath23 due to the presence of the atom @xmath24 can be calculated according to @xmath25 equation ( [ pottwo ] ) can be generalized to the @xmath26-atom potential @xmath27 .
in particular , for spherically symmetric atoms ( @xmath28 @xmath5 @xmath29 ) , one finds @xcite @xmath30,\end{gathered}\ ] ] where the symbol @xmath31 introduces symmetrization with respect to the atomic positions @xmath32 , and the corresponding force on the atom @xmath33 ( @xmath33 @xmath5 @xmath34 ) is @xmath35
let us consider a guest atom in a host medium . in this case ,
the local electromagnetic field at the position of the atom , i.e. , the field the atom interacts with is different from the macroscopic field .
this difference can be taken into account by appropriately correcting the macroscopic field to obtain the local field relevant for the atom field interaction . within macroscopic electrodynamics , a way to introduce local - field corrections is offered by the real - cavity model .
under the assumption that the guest atom ( at position @xmath1 ) is well separated from the neighboring host atoms , in the real - cavity model it is assumed that the guest atom is located at the center of a small free - space region , which in the case of an isotropic host medium is modeled by a spherical cavity of radius @xmath36 .
accordingly , the system of the unperturbed host medium of permittivity @xmath37 and permeability @xmath38 is replaced with the system whose permittivity @xmath39 and permeability @xmath40 read @xmath41 the cavity radius @xmath36 is a model parameter representing an average distance from the atom to the nearest neighboring atoms constituting the host medium ; it has to be determined from other ( preferably microscopic ) calculations or experiments .
note that on the length scale of of the order of magnitude of @xmath36 , the unperturbed host medium can be regarded as being homogeneous from the point of view of macroscopic electrodynamics , i.e. , @xmath42 \mu(\mathbf{r},\omega ) = \mu(\mathbf{r}_a,\omega ) \end{array } \right\ } \
\mathrm{for}\ @xmath43 being a small positive number . in the real - cavity model , the local electromagnetic field the guest atom interacts with is the field that obeys the macroscopic maxwell equations , with the permittivity @xmath39 and permeability @xmath40 as given by eq .
( [ loc ] ) . hence , when calculating the vdw potential of the atom by means of eq .
( [ potone ] ) , the green tensor therein is obviously the green tensor of the macroscopic maxwell equations with @xmath39 and @xmath40 , respectively , in place of @xmath37 and @xmath38 , i.e , the green tensor for the electromagnetic field in the medium disturbed by a cavity - like , small free - space inhomogeneity . clearly , when considering the mutual vdw interaction of two or more than two guest atoms , each atom has to be thought of as being located at the center of a small free - space cavity , and the green tensors in eqs .
( [ pottwo ] ) and ( [ many ] ) are the green tensors of the macroscopic maxwell equations for the electromagnetic field in the medium disturbed by the corresponding cavity - like , small free - space inhomogeneities .
the task now consists in determining such a green tensor .
the electromagnetic field inside and outside each cavity exactly solves the macroscopic maxwell equations , together with the standard boundary conditions at the surface of the cavity . for an arbitrary current distribution @xmath44 , the electric field @xmath45 , for instance ,
can be given by means of the green tensor @xmath46 according to the relation @xmath47 we begin with the case where a single guest atom is embedded in a host medium of finite size and arbitrary shape , and , at a first stage , we assume that the host medium outside the cavity which contains the guest atom is homogeneous an assumption that will be dropped later . according to eq .
( [ electric ] ) , the scattering part of the green tensor , @xmath48 , describing the electromagnetic field reaching the point @xmath1 where it has originated can be decomposed into three parts , @xmath49 ) , embedded in a ( isotropic ) host medium of arbitrary size and shape .
simple examples of the three different kinds of processes are sketched .
the corresponding green tensors are @xmath50 , accounting for all scattering processes within the cavity , @xmath51 , describing ( multiple ) transmission through the cavity surface and accounting for all scattering processes in the host medium excluding reflection at the outer boundary of the cavity , and @xmath52 , including only the scattering processes involving ( multiple ) reflection at the outer boundary of the cavity . ]
where @xmath50 accounts for ( multiple ) scattering at the inner surface of the cavity , @xmath51 describes ( multiple ) transmission through the surface of the cavity without scattering at the outer surface of the cavity , and @xmath52 is associated with ( multiple ) transmission through the surface of the cavity via scattering at the outer surface of the cavity . the decomposition is schematically illustrated in fig .
[ single ] , showing a simple example each .
the term @xmath50 in eq .
( [ eq15 ] ) is nothing but the scattering part of the green tensor at the center of an empty sphere surrounded by an infinitely extended medium . in the case of a magnetoelectric medium of permittivity @xmath53 and permeability @xmath54
, it reads @xcite @xmath55 where @xmath56 ' -\varepsilon(\omega ) h_1^{(1)}(z)\left[z_0 h_1^{(1)}(z_0)\right ] ' } { \varepsilon(\omega ) h_1^{(1)}(z)\left[z_0 j_1(z_0)\right]'-j_1(z_0 ) \left[zh_1^{(1)}(z)\right]'}\ ] ] [ @xmath57 @xmath5 @xmath58 , @xmath59 @xmath5 @xmath60 , @xmath61 @xmath5 @xmath62 , the primes indicate derivatives ] , with @xmath63 and @xmath64 , respectively , being the first spherical bessel function and the first spherical hankel function of the first kind , @xmath65 at this point it should already be mentioned that since the vdw potential depends on the green tensor at all frequencies , an expansion for small @xmath66 of the form @xmath67- 3\,\varepsilon(\omega)-1}{[2\varepsilon(\omega)+1]^2}\ , \frac{c}{i\omega
r_\mathrm{c}}\\ + 9\,\frac{\varepsilon(\omega)n^3(\omega)}{[2\varepsilon(\omega)+1]^2 } -1+o\left(\frac{\omega r_\mathrm{c}}{c}\right ) \ ] ] and neglect of the term @xmath68 , as it is done in ref .
@xcite for the calculation of the decay rate of an excited atom , has to be applied with care ( see sec .
[ subsub_nonlin ] ) . following the arguments given in ref .
@xcite , the term @xmath51 in eq .
( [ eq15 ] ) can always be written in the form @xmath69 where @xmath70 is the scattering part of the green tensor of the undisturbed host medium ( i.e. , the medium without the cavity ) , and the global factor @xmath71 can simply be determined from examining the green tensor @xmath72 which links a source at the center of the cavity with the transmitted field at a point @xmath73 in the medium outside the cavity in the case of an infinitely extended medium .
for a magnetoelectric medium , the green tensor @xmath74 reads @xcite @xmath75 e^{iq}\end{gathered}\ ] ] [ @xmath76 @xmath5 @xmath77 , @xmath78 @xmath5 @xmath79 @xmath80 @xmath81 @xmath7 @xmath82 , @xmath83 @xmath5 @xmath79 @xmath80 @xmath84 @xmath7 @xmath85 , @xmath86 @xmath5 @xmath87|\mathbf{r}$ ] @xmath7 @xmath88 , @xmath89 @xmath5 @xmath90 @xmath7 @xmath91 @xmath7 @xmath92 , where @xmath93 ' -\left[z_0j_1(z_0)\right]'h_1^{(1)}(z_0 ) } { \mu(\omega)\bigl\{j_1(z_0)\left[z h_1^{(1)}(z)\right ] ' -\varepsilon(\omega ) \left[z_0j_1(z_0)\right]'h_1^{(1)}(z ) \bigr\}}.\end{gathered}\ ] ] comparing eq .
( [ greenout ] ) with the green tensor for an undisturbed , infinitely extended host medium , @xmath94 e^{iq},\ ] ] we see that the factor @xmath71 is equal to @xmath95 , @xmath96 inspection of eq .
( [ koeff ] ) for small @xmath97 shows that @xmath98 note that in leading order , the factor @xmath95 is the same as that found for a purely electric system ( cf .
@xcite ) . as shown in ref .
@xcite , the term @xmath52 in eq .
( [ eq15 ] ) behaves like @xmath99 for a sufficiently small cavity in a dielectric medium .
the same asymptotic behavior is observed for magnetoelectric media , since , for sufficiently small @xmath36 , the leading - order contribution to @xmath52 is determined by the electric properties . recalling that in the real - cavity model the cavity radius is assumed to be small compared with the relevant atomic and medium wavelengths ,
@xmath52 is in any case small compared with @xmath50 and @xmath51 and can thus be neglected , so that eq .
( [ eq15 ] ) [ together with eqs .
( [ greencavity ] ) , ( [ combined ] ) , and ( [ eq24 ] ) ] takes the form @xmath100 so far we have concentrated on homogeneous host media of arbitrary shapes .
an inhomogeneous medium can always be divided into a ( small ) homogeneous part that contains the cavity , such that the condition ( [ realcond ] ) is satisfied , and an inhomogeneous rest . since all the scattering processes resulting from the inhomogeneous rest can be thought of as being taken into account by the scattering part of the green tensor of the host medium , eq .
( [ greenzerleg ] ) still holds in this more general case .
note that a homogeneous medium of finite size can already be regarded as a special case of an inhomogeneous system .
hence , @xmath53 and @xmath54 , respectively , can be safely replaced by @xmath101 @xmath102 @xmath103 and @xmath104 @xmath102 @xmath105 in the equations given above [ @xmath106 , @xmath107 .
next we consider two atoms within a host medium of arbitrary size and shape , both located at the center of a small cavity , such that eqs .
( [ loc ] ) and ( [ realcond ] ) hold . in this case
, there are two contributions to the green tensor @xmath21 that determines the electromagnetic field at the position @xmath1 of atom @xmath0 , originating at @xmath108 where is atom @xmath19 located , @xmath109 here , @xmath110 accounts for ( multiple ) transmission through the cavity surfaces and ( multiple ) scattering at the inhomogeneities of the medium , excluding scattering at the outer boundaries of the cavities . as in the single - atom case ,
processes that involve ( multiple ) scattering at the outer surfaces of the cavities , described by @xmath111 , are neglected .
the situation is sketched in fig .
[ two ] . and
@xmath112 account for ( multiple ) transmission through the cavity surfaces and ( multiple ) scattering at the inhomogeneities of the host medium including ( multiple ) scattering at the outer boundaries of the cavities , respectively .
] by means of the same arguments as in the single - atom case , we find a global factor that accounts for transmission through the surfaces of the cavities such that the field at @xmath73 outside the cavities which results from @xmath108 , may be calculated from @xmath113 . due to the property @xmath46 @xmath5 @xmath114 of the green tensor , we also find that @xmath115 determines the electromagnetic field at point @xmath1 originating at point @xmath73 outside the cavities .
we may thus write @xmath21 in the form @xmath116
within the framework of the real - cavity model , local - field corrected vdw interactions can now be calculated by combining the formulas given in sec .
[ sec_basic ] and [ sec_real ] , where we again begin with the case of a single guest atom .
we substitute @xmath117 as given by eq .
( [ greenzerleg ] ) into eq .
( [ potone ] ) to obtain @xmath118 where @xmath119\end{aligned}\ ] ] [ @xmath120 and @xmath121 according to eqs .
( [ greencavity1 ] ) and ( [ koeff ] ) , respectively ] . in order to make contact with a microscopic description and provide a deeper insight into the validity of the real - cavity model , let us first restrict our attention to to weakly polarizable and/or magnetizable host material so that the linear approximation applies . in this case , we may regard @xmath37 and @xmath38 as being only slightly perturbed from their free - space values , @xmath122 [ @xmath123 @xmath124 @xmath125 , @xmath126 @xmath124 @xmath125 ] , and approximate @xmath128 , eq .
( [ u1 ] ) , by @xmath129,\end{gathered}\ ] ] where we have assumed a spherically symmetric atom . by straightforward calculation , one can then show that @xmath130 where @xmath131 is the number density of the medium atoms , @xmath132 ( @xmath1 @xmath133 @xmath134 ) is the vdw potential of two unpolarized , neutral ( ground state ) atoms in free space , with the electric and magnetic parts being given by @xcite @xmath135\\ = -\frac{\hbar}{32\pi^3\varepsilon_0 ^ 2 l^6 } \int _
0^\infty \mathrm{d}u\ , \alpha_a(iu)\alpha(iu)g(ul / c)\end{gathered}\ ] ] and @xmath136 \!\cdot\!\bm{\nabla}_\mathbf{s}\!\times\ ! { \mbox{\textbf{\textsf{g}}}}_\mathrm{free}(\mathbf{s},\mathbf{r}_a , iu)\right\ } _ { \mathbf{s}=\mathbf{r } } \\
= \frac{\hbar\mu_0 ^ 2 } { 32\pi^3l^4 } \int
_ 0^\infty \mathrm{d}u\ , u^2\alpha_a(iu)\beta(iu)h(ul / c)\end{gathered}\ ] ] [ @xmath137 , @xmath138 , @xmath139 ; @xmath140 , free - space green tensor ; @xmath141 , polarizability ; @xmath142 , magnetizability of the medium atoms ] , where we have made the identifications @xmath143 to find the linear approximation to @xmath144 , eq .
( [ u2 ] ) , let us first consider @xmath145 applying the linear born expansion ( see , e.g. , ref .
@xcite ) @xmath146 with @xmath147 @xmath148\!\cdot\!\bm{\nabla}_s \!\times\!{\mbox{\textbf{\textsf{g}}}}_\mathrm{free}(\mathbf{s},\mathbf{r}',\omega)\end{gathered}\ ] ] to the calculation of the green functions @xmath149 and @xmath150 in eq .
( [ eq41 ] ) , we derive @xmath151\!\cdot\!\bm{\nabla}_s \!\times\!{\mbox{\textbf{\textsf{g}}}}_\mathrm{free}(\mathbf{s},\mathbf{r}',\omega ) \\
-\frac{\omega^2\chi(\mathbf{r}_a,\omega ) } { c^2 } \int \mathrm{d}^3s\ , { \mbox{\textbf{\textsf{g}}}}_\mathrm{free}(\mathbf{r}_a,\mathbf{s},\omega)\!\cdot\ ! { \mbox{\textbf{\textsf{g}}}}_\mathrm{free}(\mathbf{s},\mathbf{r}_a,\omega)\\ + \,\zeta(\mathbf{r}_a,\omega ) \int \mathrm{d}^3s \left[{\mbox{\textbf{\textsf{g}}}}_\mathrm{free}(\mathbf{r},\mathbf{s},\omega ) \!\times\!\overleftarrow{\bm{\nabla}}_s\right]\!\cdot\!\bm{\nabla}_s \!\times\!{\mbox{\textbf{\textsf{g}}}}_\mathrm{free}(\mathbf{s},\mathbf{r}',\omega)\end{gathered}\ ] ] ( @xmath152 , volume of the unperturbed host medium ) . equation ( [ greenborn ] ) reveals that , in contrast to @xmath153 , @xmath70 has no zero - order contribution with respect to @xmath154 and @xmath155 .
thus , the linear approximation to @xmath144 can be obtained by substituting eq .
( [ greenborn ] ) into eq .
( [ u2 ] ) , letting @xmath153 @xmath5 @xmath125 , and applying eqs . ( [ chi ] ) and ( [ zeta ] ) , which results in @xmath156 as expected ,
the overall potential in linear approximation is simply @xmath157 ( @xmath158 , volume of the host medium excluding the volume of the cavity ) , where we have recalled eq .
( [ realcond ] ) .
the results found in linear approximation , particularly eq .
( [ eq44 ] ) , indicate that @xmath36 can indeed be identified with an average inter - atomic distance .
hence , the real - cavity model can be regarded as being valid down to a microscopic scale .
since @xmath128 can be described as the sum of the vdw interactions between the guest atom and all the atoms in an infinitely extended homogeneous medium , it does not lead to a force but to a more or less significant shift of the overall potential , which is dominated by the interaction of the guest atom with the neighboring host atoms .
note that we have demonstrated these results up to linear order in @xmath154 and @xmath155 . in principle , a generalization beyond the linear born expansion should be possible , although associated with extensive calculations . in this case
many - atom vdw interactions have to be taken into account .
let us return to the integrals in eqs .
( [ u1 ] ) and ( [ u2 ] ) . to find the local - field correction
, one has in fact to consider the leading - order terms in @xmath36 after the frequency integrals have been performed .
however , since @xmath159 , @xmath160 , and @xmath161 provide cut - off functions for the @xmath162-integrals [ giving their main contribution for @xmath163 , with @xmath164 being the maximum of all relevant atomic and medium resonance frequencies ] and assuming that @xmath165 , we may expand @xmath120 and @xmath121 in the integrands , keep the leading - order terms in @xmath166 , and integrate over @xmath162 afterwards . this procedure can be justified by the good convergence behavior in the asymptotic limit of small @xmath36 , which we have confirmed numerically .
thus , using eq .
( [ cexp ] ) , we can write @xmath128 , eq .
( [ u1 ] ) , in the form @xmath167 + 3\varepsilon_a(iu)+1}{5\left[2\varepsilon_a(iu)+1\right]^2}\ , \frac{u^2 } { r_\mathrm{c}}\right]\\ \times\mathrm{tr}\,\bm{\alpha}_a(iu).\end{gathered}\ ] ] as already mentioned , @xmath128 can be ignored if one is only interested in the force on the atom ( see also app .
[ app1 ] ) .
similarly , using eq .
( [ d ] ) , we can write @xmath144 , eq .
( [ u2 ] ) , as @xmath168 ^ 2\\ \times\mathrm{tr}\left[\bm{\alpha}_a(iu)\!\cdot\ ! { \mbox{\textbf{\textsf{g}}}}_\mathrm{med}^{(1)}(\mathbf{r}_a,\mathbf{r}_a , iu)\right].\end{gathered}\ ] ] with eq . ( [ forceone ] ) and the condition ( [ realcond ] ) , one obtains @xmath169 ^ 2\nonumber\\ & \times\bm{\nabla}_{\!\!a } \mathrm{tr}\left[\bm{\alpha}_a(iu)\cdot { \mbox{\textbf{\textsf{g}}}}_\mathrm{med}^{(1)}(\mathbf{r}_a,\mathbf{r}_a , iu)\right]\end{aligned}\ ] ] for the local - field corrected vdw force on a ground - state atom within an arbitrary magnetoelectric system .
in the limit of the atom being embedded in a very dilute medium , @xmath170 @xmath124 @xmath125 and @xmath171 @xmath124 @xmath125 , the local - field correction becomes negligible and eqs .
( [ ucapprox])([forceloc ] ) approach the uncorrected result [ eqs .
( [ potone ] ) and ( [ forceone ] ) ] , as expected . as a note of caution ,
we remark that , quite contrary , in the limit @xmath36 @xmath172 @xmath173 , the local - field corrected result does not approach the uncorrected one , which might be obtained when the atom is thought of as not being placed in an empty free - space region , but being placed in a region which is continuously filled with medium .
this is due to the fact that there is no continuous @xmath36 @xmath172 @xmath173 transition from the corrected setup ( an atom inside an empty free - space cavity surrounded by a medium ) and the uncorrected one ( an atom inside the medium ) : even as the cavity radius is made smaller and smaller , the atom will always remain inside a free - space region .
equation ( [ forceloc ] ) shows that , in comparison to the uncorrected vdw force ( [ forceone ] ) , the contributions to the are enhanced by a factor between @xmath174 and @xmath175 owing to the local - field correction .
note that the enhancement factor varies with @xmath162 and the uncorrected contributions can in general ( in particular , for magnetoelectric media ) have different signs for different values of @xmath162 .
hence , the local - field corrected vdw force can be enhanced or reduced .
for example , for ( purely electric ) media important in biological applications , local - field corrections can lead to enhancements by factors of @xmath176 [ for water with @xmath177 or @xmath178 ( for typical plasma membrane compositions with @xmath179 @xcite ) .
using eqs .
( [ pottwo ] ) and ( [ greenzerleg3 ] ) , the local - field corrected two - atom vdw potential reads @xmath180 .
\\[-2ex]\end{gathered}\ ] ] in analogy to sec .
[ subsub_nonlin ] , we may expand the coefficients @xmath181 and @xmath182 to leading order in @xmath166 , as given by eq .
( [ d ] ) , and integrate afterwards .
hence , eq . ( [ u_ab ] ) leads to @xmath183 ^ 2 \left[\frac{3\varepsilon_b(iu)}{2\varepsilon_b(iu)+1}\right]^2 \\\times\ ! \mathrm{tr}[\bm{\alpha}_a(iu)\!\cdot\ ! { \mbox{\textbf{\textsf{g}}}}_\mathrm{med}(\mathbf{r}_a,\mathbf{r}_b , iu ) \!\cdot\!\bm{\alpha}_b(iu)\!\cdot\ ! { \mbox{\textbf{\textsf{g}}}}_\mathrm{med}(\mathbf{r}_b,\mathbf{r}_a , iu ) ] .
\\[-2ex]\end{gathered}\ ] ] in comparison to the uncorrected potential ( [ pottwo ] ) , the contributions to the @xmath162-integral are enhanced by a factor which varies between @xmath174 and @xmath184 , so that altogether an enhancement or an reduction of the local - field corrected two - atom force may be observed ( recall the remark at the end of sec .
[ subsub_nonlin ] ) .
equation ( [ uabl ] ) can easily be extended to the mutual interaction of @xmath26 atoms .
substituting eqs .
( [ d ] ) and ( [ greenzerleg3 ] ) into eq .
( [ many ] ) , one finds that , for isotropic atoms , the local - field corrected @xmath26-atom potential is given by @xmath185 ^ 2\cdots \left[\frac{3\varepsilon_n(iu)}{2\varepsilon_n(iu)+1}\right]^2 \alpha_1(iu)\cdots\alpha_n(iu)\\ \times
\mathrm{tr}\left[{\mbox{\textbf{\textsf{g}}}}_\mathrm{med}(\mathbf{r}_1,\mathbf{r}_2,iu ) \cdots{\mbox{\textbf{\textsf{g}}}}_\mathrm{med}(\mathbf{r}_n,\mathbf{r}_1,iu)\right].\end{gathered}\ ] ] as a simple example , let us consider two isotropic ground - state atoms embedded in magnetoelectric bulk material characterized by @xmath53 and @xmath54 . substituting eq .
( [ greenbulk ] ) into eq .
( [ uabl ] ) , we obtain @xmath186 ^ 4 \left\{3 + 6n(iu)ul / c+5[n(iu)ul / c]^2\right.\\ \left.+2[n(iu)ul / c]^3+[n(iu)ul / c]^4\right\}e^{-2n(iu)ul / c},\end{gathered}\ ] ] where @xmath187 @xmath188 @xmath7 @xmath189 . in the retarded limit , @xmath190 @xmath191 @xmath192 ( with @xmath193
denoting the minimum of all the relevant atomic and medium resonance frequencies ) , the main contributions to the @xmath162-integral come from the range where @xmath194 with this inserted , eq .
( [ twobulk ] ) simplifies to @xmath195 where @xmath196 ^ 4.\ ] ] in the non - retarded limit , , we may set @xmath197 @xmath198 @xmath125 and approximate the term in the curly brackets by @xmath199 . in this way , eq .
( [ twobulk ] ) reduces to @xmath200 where @xmath201 ^ 4.\ ] ] equations ( [ eq57 ] ) and ( [ eq59 ] ) generalize the well - known results for the uncorrected mutual vdw interaction of two atoms in bulk material @xcite .
we see that the local - field corrected potential follows the usual @xmath202 and @xmath203 dependence in the retarded and non - retarded limits , respectively .
however , due to the influence of the local - field correction , the potential / force can be noticeably enhanced in dense media .
in this paper , we have studied the local - field corrected vdw interaction of ground - state atoms embedded in a finite magnetoelectric medium of arbitrary size and shape .
we have started from the familiar expressions for the one- and many - atom vdw potentials in leading - order perturbation theory , given in terms of the associated ( classical ) green tensors and the atomic polarizabilities . employing the real - cavity model for the local - field correction and using the methods developed in ref .
@xcite , we have derived general expressions for the required local - field corrected green tensors . using these results ,
we have obtained geometry - independent , general formulas for the local - field corrected one- and many - atom ( ground - state ) vdw potentials and the corresponding forces .
the derivation is based on the assumption that the radius of the free - space cavity , in which a guest atom is thought of as being located and which is a measure of the average distance of the guest atom from the surrounding host atoms , is small compared to the other characteristic lengths of the problem .
that is to say , it should be small compared to the distance of a guest atom from surfaces of discontinuity of the host medium , the distances between the guest atoms in the case of many - atom interactions , and the wavelengths associated with the electromagnetic response of the guest atoms and the host medium .
it should be pointed out that under these conditions the forces do not explicitly depend on the cavity radius .
we have found that the local - field correction leads to a position - independent shift of the single - atom potential , while at the same time enhancing the contributions to the position - dependent part by a frequency - dependent factor that may vary between @xmath174 and @xmath175 .
the frequency - dependent contributions to the two - atom potential , on the other hand , are enhanced by a factor between @xmath174 and @xmath184 owing to the local - field correction .
the contributions to the general @xmath26-atom potential are enhanced accordingly . in all cases ,
the respective vdw force can be enhanced or reduced . as a simple example
, we have calculated the mutual vdw potential of two ground - state atoms placed within a magnetoelectric bulk medium .
the result shows that the potential follows the well - known asymptotic @xmath202 and @xmath203 power laws in the retarded and non - retarded regimes , respectively , where the local - field correction leads to an enhancement of the associated proportionality constants .
we conclude by noting that , besides vdw interaction and spontaneous decay , the local - field corrected green tensors obtained can also be used to account for the effects of the local field in a number of important processes ( e.g. , fluorescence , light scattering , energy transfer , etc . ) .
evidently , one can expect that these processes are affected by local - field corrections in a similar way as found here .
this work was supported by the deutsche forschungsgemeinschaft and by the ministry of science of the republic of croatia .
, s.y.b . , and d .- g.w .
are grateful to ho trung dung for discussions .
in sec . [ sec_local ] , we have encountered the potential term @xmath128 , eq .
( [ u1 ] ) , which is equal to the potential of an atom at the center of a free - space cavity embedded in a bulk medium of permittivity @xmath101 and permeability @xmath104 .
more generally , the potential of an atom at a small displacement @xmath204 from the center of the cavity is given by eq .
( [ potone ] ) with the green tensor @xcite @xmath205'^2}{\xi^2_a } + c^m_l(\omega)j^2_l(\xi_a)\right]\\ \times(\mathbf{e}_{\theta } \mathbf{e}_{\theta}+\mathbf{e}_{\phi}\mathbf{e}_{\phi})\bigr\}.\end{gathered}\ ] ] ( @xmath206 , @xmath207 , @xmath208 , spherical unit vectors associated with @xmath204 ; @xmath209 @xmath5 @xmath210 ) . here
@xmath211 is given by eq .
( [ greencavity1 ] ) , with the subscript @xmath212 , and @xmath213 @xmath5 @xmath214 $ ] . from @xmath215
we have @xmath216 ^ 2\\ \times\bigl[l(l+1)\mathbf{e}_{r}\mathbf{e}_{r}+ { \textstyle\frac{1}{2}}(l+1)^2(\mathbf{e}_{\theta}\mathbf{e}_{\theta}+ \mathbf{e}_{\phi}\mathbf{e}_{\phi})\bigr]\end{gathered}\ ] ] for the small @xmath209 expansion of the green tensor .
as seen , in the limit @xmath217 , only the @xmath190 @xmath5 @xmath125 term survives , leading to the potential ( [ u1 ] ) .
the force on an atom close to the cavity center is determined by the derivative of the green function with respect to @xmath218 .
therefore , for small displacements of the atom from the cavity center , it is given by the @xmath190 @xmath5 @xmath219 term in the above expansion , and we find the linear force @xmath220 with @xmath221 [ recall eq .
( [ greencavity1 ] ) ] , which vanishes at the cavity center . for small cavity radii eq .
( [ k ] ) reduces to @xmath222 -6\varepsilon(iu)-4 } { [ 3\varepsilon(iu)+2]^2}\,\frac{c^3 } { u^3 r_\mathrm{c}^3}\right\},\end{gathered}\ ] ] leading to a force pointing away from the center of the cavity ( corresponding to an unstable equilibrium ) for materials with purely dielectric properties and to a harmonic force for purely magnetic materials . | based on macroscopic quantum electrodynamics in linearly and causally responding media , we study the local - field corrected van der waals potentials and forces for unpolarized ground - state atoms placed within a magnetoelectric medium of arbitrary size and shape .
we start from general expressions for the van der waals potentials in terms of the ( classical ) green tensor of the electromagnetic field and the atomic polarizability and incorporate the local - field correction by means of the real - cavity model . in this context
, special emphasis is given to the decomposition of the green tensor into a medium part multiplied by a global local - field correction factor and , in the single - atom case , a part that only depends on the cavity characteristics .
the result is used to derive general formulas for the local - field corrected van der waals potentials and forces . as an application
, we calculate the van der waals potential between two ground - state atoms placed within magnetoelectric bulk material . |
one subject that has received considerable attention in recent years is the return to equilibrium of systems arising from statistical mechanics .
one of the simplest models of interest is given by the kinetic fokker - planck equation _ t _ t = _ t,_0 = , = 1 2 _ p^2 - p _
p + p _ q - v(q)_p , p , q ^n .
this equation describes the evolution of an observable @xmath5 under the hamiltonian dynamic for the energy @xmath6 , perturbed by friction and noise : dp = -v(q ) dt -p dt + dw(t),dq = pdt . the relation between and is given by the fact that the function @xmath7 satisfies the partial differential equation , provided that the pair @xmath8 solves .
it can easily be checked by inspection that if @xmath9 is sufficiently smooth and coercive , the measure @xmath10 is invariant under this dynamic in the sense that if @xmath11 satisfies , then @xmath12 does not depend on @xmath13 .
this can be rephrased as saying that if @xmath14 is a random variable with law @xmath15 independent of the driving noise @xmath16 , then the law of @xmath17 is given by @xmath15 for all times .
underpinning much of the analysis of @xmath18 on the space @xmath19 is the guiding principle that it related to the corresponding _
witten laplacian _
@xcite _ v = - _ q + |v(q)|^2 - ( v)(q ) .
in particular , it was conjectured by helffer and nier ( * ? ? ? * conjecture 1.2 ) that @xmath18 has compact resolvent on @xmath20 if and only if @xmath21 has compact resolvent on the _ flat _ space @xmath22 .
this conjecture has been partially solved in @xcite ( see also @xcite ) in the sense that one can exhibit a large class of potentials for which it holds ( loosely , @xmath9 should grow in a sufficiently regular way at infinity ) .
recently , upper as well as lower bounds on the spectral gap of @xmath18 have been obtained in @xcite for potentials @xmath9 that satisfy certain growth conditions at infinity .
all of the results cited above make heavy use of the following two key facts : there is an explicit formula for the invariant measure of .
the friction term @xmath23 acts on all ( physical ) degrees of freedom of the system .
both of these facts are violated in the following very popular model for heat conduction .
take a finite collection of @xmath24 anharmonic oscillators with nearest - neighbor couplings , that is a classical hamiltonian system with hamiltonian given by h(p , q ) = _ i=0^n ( p_i^2 2 + v_1(q_i ) ) + _
i=1^n v_2(q_i - q_i-1 ) . here , the potential @xmath25 is the interaction potential between neighboring oscillators , whereas @xmath26 is a _ pinning potential_. this system is then put in contact with two heat baths at _ different _ temperatures @xmath27 and @xmath28 .
we model the interaction with the heat baths by the standard langevin dynamics , so that the equations of motion of our system are given by dp_0 & = - _ 0
p_0dt - v_1(q_0)dt - v_2(q_0 - q_1)dt + _ 0dw_0 + dp_i & = -v_1(q_i)dt - v_2(q_i - q_i-1)dt - v_2(q_i - q_i+1)dt [ e : main ] + dp_n & = - _ n p_ndt - v_1(q_n)dt - v_2(q_n - q_n-1)dt + _ ndw_n + dq_j & = p_jdt . here , we set @xmath29 , the index @xmath30 runs from @xmath31 to @xmath32 and the index @xmath33 runs from @xmath34 to @xmath35 . as described in @xcite , the rigorous analysis of this model and in particular the derivation of fourier s law ( or the proof of its breakdown ) is an outstanding mathematical problem of great interest to the applied community . if @xmath36 , then one can check as before that this set of equations has a unique invariant measure , which has density @xmath37 with respect to the lebesgue measure , where @xmath38 is the hamiltonian from . when the two temperatures @xmath27 and @xmath28 are different however , much less is known .
in particular , as we will see immediately , even the _ existence _ of an invariant probability measure is an open problem in some cases as simple as @xmath39 and @xmath40 .
this model has been the subject of many studies , both from a numerical and from a theoretical perspective .
the purely harmonic case has been solved explicitly in @xcite .
though no explicit solution is known in the anharmonic case , a wealth of numerical experiments exist , see for example @xcite and references therein .
since we will mainly focus on the theoretical aspects of the model , we refer to @xcite , which seems to be one of the first rigorous studies of the anharmonic case .
it was shown in @xcite that if @xmath41 and @xmath42 behave approximately like @xmath43 and @xmath44 respectively at infinity then , provided that @xmath45 , there exists a unique invariant measure for .
the statement that was proved was actually stronger than that , namely it was shown that the generator @xmath18 of has compact resolvent in every space of the form @xmath46 with @xmath47 . in @xcite , it was also shown , using entirely probabilistic rather than functional - analytic techniques , that the condition @xmath48 is sufficient for the existence and uniqueness of an invariant measure for .
furthermore , the compactness of the corresponding semigroup in some weighted @xmath49 space was also proved there .
this left open the case @xmath50 which is the subject of the present study .
to our knowledge , no previous rigorous results exist in this case , although some interesting theory has been developed recently in @xcite . at first sight
, one might think that there is no _ a priori _ reason for the behavior of to differ in any essential way from the case @xmath51 where spectral gap results are known .
such wishful thinking turns out to be overly optimistic . even in the simplest possible scenario , that is when
@xmath39 and @xmath40 , we will show in theorem [ theo : threeosc ] below that the compactness property of the resolvent of @xmath18 is destroyed as soon as @xmath52 .
furthermore , when @xmath53 , it will be shown in theorem [ theo : compactchain ] that the essential spectrum of @xmath18 ( always in a weighted @xmath54 space of the type considered before ) extends all the way to @xmath34 .
these negative results hold even in the case where @xmath36 , showing that having the friction acting on all physical degrees of freedom is a crucial assumption for the helffer - nier conjecture to hold .
the reason why the behavior of changes so drastically when @xmath50 can be understood heuristically by the appearance of _ breathers _ ( see for example @xcite ) .
breathers are dynamically stable , spatially localized , periodic orbits that arise in the noise - free translation invariant ( @xmath55 ) version of .
good approximations to these orbits persist in , especially if @xmath35 is large .
it is therefore possible to put the system into a state where most of its energy is localized in a few oscillators located in the middle of the chain .
on the other hand , energy can be dissipated only through the terms @xmath56 appearing in the equations for the first and the last oscillator .
therefore , one expects the energy of the system to decay extremely slowly .
the appearance of breathers can be proved in the case where the strength of the nearest - neighbor coupling is much weaker than the strength of the pinning potential . at high energies ,
this is precisely the case when @xmath50 .
this discussion shows that even the mere _ existence _ of an invariant measure is a non - trivial problem in this model unless @xmath57 , in which case one can check explicitly that the usual boltzmann - gibbs distribution is invariant .
this differs from many other systems in that the _ existence _ of an invariant measure , not its _ uniqueness _
, is difficult .
in fact , the uniqueness of an invariant measure for a chain of arbitrary length follows quickly from the hypoellipticity of the generator and the hamiltonian structure once the existence of an invariant measure is established .
we are at the moment unable to provide a general proof that shows the existence of an invariant measure for a chain of arbitrary length . however , in the case of a chain comprising of @xmath58 oscillators , we show in theorem [ theo : wholething ] below that there exists a unique invariant measure , provided that the coupling potential is harmonic and the pinning potential is homogeneous of sufficiently high degree .
this result is proven by first obtaining an effective dynamics when most of the energy is concentrated in the central oscillator .
this effective dynamics is then used to construct a lyapunov function whose level sets are compact and hence , by a variation on the classical kryloff - bogoliouboff method , implies the existence of an invariant measure .
the remainder of this article is organized as follows . in section [ sec : formalcalc ] , we give a formal calculation that shows how it is possible to relate the spectral properties of @xmath18 to the scaling properties of the potentials @xmath26 and @xmath25 . the results given by these formal calculations are then compared to numerical simulations .
we proceed in section [ sec : negative ] to the proof of the _ negative _ results concerning the lack of compactness and/or of a spectral gap for @xmath18 . in section [ sec : effective ] we derive effective equations of motion for the system of three oscillators in the regime when a breather is present . here , we make heavy use of the compensator techniques from averaging / homogenization theory @xcite .
finally , we show in section [ sec : existence ] that in the simplest case of a chain of three oscillators with harmonic coupling potentials , one can show the existence of an invariant measure for any degree of homogeneity greater than @xmath59 for the pinning potential .
perhaps more interesting then this last result is the method of proof .
we derive as system of effective equations and prove their accuracy that when the energy of middle oscillator is large .
these effective equations give insight into the mechanism of energy dissipation and are used to construct a lyapunov function .
we would like to thank the university of bonn ( and the then - nascent hausdorff institute ) for providing the pleasant working environment where this project began .
mh would like to thank jean - pierre eckmann and luc rey - bellet for many long discussions about this problem .
we would also like to thank arnaud guillin for his encouragements which aided in completing this text in a ( more or less ) timely manner .
mh acknowledges support from epsrc fellowship ep / d071593/1 .
jcm acknowledges the support of the sloan foundation and national science foundation through a pecase award ( dms 04 - 49910 ) and a standard award ( dms 06 - 16710 ) .
in this section , we first perform a formal calculation that allows us to get a feeling of the speed at which energy is extracted from such a system .
consider the simplest non - trivial case , that is when @xmath60 . in order to keep things simple
, we furthermore assume that the coupling potential @xmath25 is quadratic : @xmath61 and that the pinning potential is homogeneous of degree @xmath62 : @xmath63 for some real number @xmath2 .
the equations of motion for the system of interest are thus given by dp_0 & = - _ 0 p_0dt - q_0 |q_0|^2k-2dt - ( q_0 - q_1)dt + dw_0 + dp_1 & = - q_1 |q_1|^2k-2dt - ( 2q_1 - q_0 - q_2)dt + dp_2 & = - _ 2 p_2dt - q_2 |q_2|^2k-2dt - ( q_2 - q_1)dt + dw_2 + dq_j & = p_jdt .
let us first have a look at the motion of the middle oscillator by itself , at the solution of dp = -q |q|^2k-2dt , dq= pdt .
it is easy to see that this equation is invariant under the substitution q(t ) = e^12k q(e^t),p(t ) = e^1 2 p(e^t ) , with @xmath64 .
let us now denote by @xmath65 the solution to with initial condition @xmath66 , so that @xmath67 as given by is the ( unique up to a phase ) solution to at energy @xmath68 .
since the variables @xmath69 and @xmath70 are assumed to be one - dimensional , the solution @xmath65 is periodic , say with period @xmath71 .
consider now the equation for the left oscillator , into which we substitute the ( approximate ) solution to the motion of the middle oscillator that we just found : dp_0 & = - _ 0 p_0dt - q_0 |q_0|^2k-2dt + q_0dt + dw_0 + e^12kq(e^t)dt + dq_0 & = p_0dt [ e : firstosc ] .
if @xmath68 is large compared to the typical size of @xmath72 we expect that , up to lower order corrections , the solution to this equation behaves like the superposition of the solution @xmath73 to with the exogenous forcing @xmath74 and of a highly oscillatory term of the form p_0 = e^12k - p(e^t),p = q. by symmetry , the same applies to the right oscillator @xmath75 .
applying now it s formula to the total hamiltonian @xmath38 for yields d dt h(t ) = ( _ 0 t_0 + _ 2 t_2 - _ 0 p_0 ^ 2(t ) - _ 2 p_2 ^ 2(t ) ) . in light of the above discussion , we take @xmath76 for @xmath77 so @xmath78 . from its definition , observe that @xmath79 as @xmath80 for @xmath77 . furthermore ,
when the energy @xmath68 of the middle oscillator is large , we expect the product @xmath81 to average out to a small quantity when integrated over time periods much larger than @xmath82 due to the highly oscillating , mean zero @xmath83 . applying this line of reasoning to shows that , in the regime where most of the energy of the system is concentrated in the middle oscillator , one expects to have d dt h(t ) - ( _ 0 p_0 ^ 2 + _ 2 p_2 ^ 2 ) - ( _ 0 + _ 2 ) _ k h(t)^2k - 1 , where @xmath84 is the variance of the function @xmath85 introduced in , that is _ k = 1_0^p^2(s)ds - 1 ^ 2 ( _ 0^p(s)ds)^2,p(t ) = q(t ) . the dependence on @xmath86 comes from the fact that @xmath87 is the position of the free oscillator with potential @xmath88 . actually , one expects this behavior to be correct even in a chain with more than just three oscillators .
if the chain has @xmath89 oscillators and most of the energy is stored in the middle oscillator , one expects the motion of the endpoints to be given by the superposition of a _ slow _ motion and a highly oscillatory _ fast _ motion @xmath90 with a scaling of the type p_0 = e^12k - ( 2n-1 ) p(e^t ) , where @xmath91 is the ( unique ) periodic function such that @xmath92 and such that the integral of @xmath91 over one period vanishes .
( this is because we assume that the nearest - neighbor couplings are linear . )
this suggests that in the general case of a chain of length @xmath89 , the energy of the system decreases like d dt h(t ) - ( _ 0 + _ n ) _ k , n h(t)^2nk + 1 - 2n , for @xmath93 the variance of @xmath91 ( of course , @xmath94 as defined above )
. if @xmath95 is even , the worst - case scenario is obtained by storing most of the energy in one of the two middle oscillators .
since their distance to the boundary is the same as the distance of the middle oscillator to the boundary in the chain of length @xmath96 , we expect the rate of decay of the energy to be similar in both cases .
if the coupling potential is not quadratic but homogeneous of degree @xmath97 , one can still perform a calculation similar to the one we just did , but one has to be more careful .
when looking at the influence of @xmath98 on @xmath99 say , one should take into account whether the fast oscillations of @xmath98 are of order smaller or larger than @xmath31 .
if they are of order smaller than one , one can linearize the coupling potential .
if they are or order larger than one , one should multiply them by the scaling exponent arising in the coupling .
suppose as before that the chain contains @xmath89 oscillators and that most of it s energy is stored in the middle oscillator ( oscillator @xmath100 ) .
assume that the amplitudes in the fast oscillations of @xmath101 and @xmath98 scale like @xmath102 and @xmath103 respectively .
( recall that @xmath104 is the exponent giving the period of the oscillations . )
one then has @xmath105 .
the values of @xmath106 with @xmath107 are given by the following recursion formula : _ i = \
{ rl ( 2 - 1)(_i+1 - ) - & + _ i+1 - 2 & . using this formula ,
one can then compute @xmath108 . note that if @xmath109 , one obtains @xmath110 which agrees with the value for @xmath111 obtained previously .
one would expect these scaling relations to hold at high energies , even if the potentials are not exactly homogeneous .
one can then still perform most of the analysis presented below by splitting the right - hand side of the equations into a homogeneous part and a remainder term and by assuming that the remainder gets small ( in a suitable sense ) at high energies . in this section ,
we show that there is a surprisingly good agreement , even over extremely long time intervals , between numerical simulations of and the predictions and . in order to compare the two
, we introduce the function h^(k , n)(p , q ) = h^(p , q),= 2n ( 1 - 1 k ) . plugging this into
, we get the prediction h^(k , n)(t ) h^(k , n)(0 ) - ( _ 0 + _
2)n(2- ) _ k , n .
a straightforward numerical simulation ( essentially integration of the free equation ) furthermore allows to compute the values of @xmath93 to very high precision .
for example , we obtain _ 2,1 0.63546991 , _ 3,1 0.42363371 .
we performed numerical simulations for the cases @xmath112 and @xmath113 .
both simulations were performed using a modification of the classical strmer - verlet method ( see for example @xcite ) to take into account for the friction and the noise .
the modification was done in such a way that the resulting method is still of order two .
the simulation for @xmath114 was performed at a stepsize @xmath115 , and the simulation for @xmath116 was performed at a stepsize @xmath117 . both simulations used @xmath118 and @xmath119 .
( figures/)figure4 ( figures/)figure6 the prediction obtained from with the values is shown as a dashed line on these figures , but it fits the numerics so well that it is nearly invisible .
we emphasize that there were _ no _ free parameters in the fit , all constants are predicted by the theory .
note that the timescale in the second picture differs by a factor @xmath120 from the timescale in the first picture . in this section
, we argue that the energy decay rate predicted by also yields a prediction on the qualitative nature of the spectrum of the generator of in the weighted space @xmath20 , where @xmath15 is the invariant measure .
the idea is to model the behavior of the energy @xmath121 by a one - dimensional diffusion of the type dx = ( b(x)+ a(x)a(x ) ) dt + a(x ) dw(t ) .
it is well - known that the invariant measure for is given by \(dx ) = z^-1 ( 2_0^x b(t )
a^2(t)dt ) .
since we expect the invariant measure of to behave roughly like @xmath122 ( up to lower - order corrections ) , we should choose @xmath123 such that @xmath124 for large @xmath125 . combining this with
, we obtain the model b(x ) = - x^ , a(x ) = x^/2,= 2nk + 1 - 2n , which has @xmath126 as its invariant measure ( we restrict ourselves to the half - space @xmath127 ) .
note that since @xmath2 ( the pinning potential grows faster than the coupling potential ) , one has always @xmath128 . with the choice , the generator for is then given by ( f)(x ) = 12 _ x ( x^_x f)(x ) - x^ _ x f(x ) .
since the operator @xmath129 is a unitary operator from the weighted space @xmath130 to the flat @xmath54 space , the operator @xmath18 is unitarily equivalent to the operator @xmath131 on @xmath54 given by ( _ 1 f)(x ) = 12 _ x ( x^_x f)(x ) - x^-12(x - ) . at this point
, we recall that if @xmath132 is a strictly increasing differentiable function with @xmath133 and @xmath134 , then the operator ( u_f)(x ) = f((x ) ) is a unitary operator from @xmath135 to itself which furthermore satisfies the identity @xmath136 . under conjugation with @xmath137
, we see that one has the unitary equivalences _ x _ x - ( x ) 2 ( (x))^2 , v(x ) v((x ) ) . choosing @xmath138 , we see that @xmath131 is unitarily equivalent to the operator @xmath139 given by ( _ 2 f)(x ) & = 12 ( 2 - 2_x + 4x ) ( 2 - 2_x - 4x ) - x^2 - 22 - 2 ( x^2 2- - ) + & = ( 2-)^2 8_x^2 + ( ( 2- ) 16 - ^232 ) x^-2 - x^2 - 2 2 - 2 ( x^22- - ) this is a schrdinger operator with a potential that behaves at infinity like @xmath140 .
it follows that if @xmath141 , then the operator @xmath139 has compact resolvent .
if @xmath142 , the operator @xmath139 does not have compact resolvent , but it still has a spectral gap ( since one can see that its essential spectrum is the interval @xmath143 ) . if @xmath144 , then @xmath34 belongs to the essential spectrum of @xmath139
see for example @xcite for a proof .
it is then a natural conjecture that the spectrum of the generator of on the @xmath54 space weighted by the invariant measure has the same behavior ( as a function of the parameter @xmath145 ) as just described .
the next section is a step towards a proof of this conjecture .
the aim of this section is to obtain information on the location of the essential spectrum of the generator @xmath18 for .
this will be accomplished by using ideas from averaging / homogenization theory to build a set of approximate eigenvectors . since @xmath18 is not self - adjoint
, there are various possible definitions of its essential spectrum ( see @xcite or @xcite for a survey ) .
we choose to retain the following definition : for @xmath146 a closed densely defined operator on a banach space @xmath147 , the essential spectrum @xmath148 is defined as the set of all values @xmath149 such that @xmath150 is not a semi - fredholm operator .
the set @xmath148 is contained in the corresponding sets for all other alternative definitions of the essential spectrum appearing in the abovementioned works . in this sense , the statement `` @xmath151 '' used here is the strongest .
in particular , it is contained in the set _ k ( ) ( t + k ) , where @xmath152 denotes the ideal of all compact operators on @xmath147 and @xmath153 denotes the spectrum of an operator @xmath146 . we will make use of the following generalization of weyl s criterion @xcite , which gives a useful criterion for identifying the essential spectrum : [ prop : compact ] let @xmath146 be a closed densely defined operator on a banach space @xmath147 . for any @xmath154 , @xmath151 if and only if there exist sequences @xmath155 and @xmath156 of elements in @xmath157 and @xmath158 respectively with @xmath159 and having no convergent subsequence such that @xmath160 . see ( * ? ? ?
* theorem 9.1.3 ) and ( * ? ? ? * theorem iv.5.11 ) , or the original work @xcite .
there are situations in which , even though it is difficult to locate the essential spectrum precisely , one can nevertheless exhibit a sequence @xmath155 as above such that @xmath161 remains bounded . in that case , one has : [ prop : compact2 ] let @xmath146 be a closed densely defined operator on a banach space @xmath147 with non - empty resolvent set . if there exists a sequence @xmath155 of elements in @xmath157 with @xmath162 and having no convergent subsequence such that @xmath163 , then @xmath146 does not have compact resolvent .
this claim follows from the fact that the compactness of the resolvent is equivalent to the statement that sets of the form @xmath164 are precompact .
before we turn to the study of the generator , we make some remarks about the regularity and the behavior of functions that scale in a particular way .
denote by @xmath165 the liouville operator associated to the free
oscillator ( p , q ) = p^2 2 + |q|^2k2k .
the constant @xmath86 is not necessarily an integer , so that in general the function @xmath166 is not @xmath167 but only @xmath168}$ ] , where @xmath169 $ ] is the integer part of @xmath62 .
we introduce the following definition : [ def : scaling ] a function @xmath170 is said to _ scale like @xmath171 _ if it satisfies the relation @xmath172 .
one has the following elementary result : [ lem : scaling ] if @xmath173 scales like @xmath171 , then @xmath174 scales like @xmath175 , @xmath176 scales like @xmath177 , and @xmath178 scales like @xmath179 . given some fixed @xmath180
, there is a one - to - one correspondence between functions @xmath181 that scale like @xmath182 and functions on the circle @xmath183 in the following way .
define a function @xmath184 by the unique positive solution to r^2 ( ) ^2 2 + r^2k ( ) ||^2k 2k = 1 , and set ( ) ( ) = ( r
( ) , r ( ) ) .
the function @xmath185 is bounded away from @xmath34 ( it is actually between @xmath186 and @xmath187 ) and , by the implicit functions theorem , it is of class @xmath168}$ ] .
a straightforward but slightly tedious calculation shows that , via @xmath188 , the operator @xmath189 is conjugated to the differential operator ( x _ f ) ( ) = ( ) f ( ) , ( ) = ^2 + r^2k-2 ( ) ||^2k .
a function @xmath190 that scales like @xmath182 is said to average out to @xmath34 if _
0 ^ 2 ( ) ( ) ( ) d= 0 . with these preliminaries
, the following result is now straightforward : [ prop : scale ] let @xmath191 scale like @xmath171 and average out to @xmath34 . then , there exists a unique solution @xmath181 to the equation x _ = , such that @xmath181 also averages out to @xmath34 .
furthermore , @xmath181 scales like @xmath192 and one has @xmath193 with @xmath194 , r+1\}$ ] .
let @xmath181 be the unique function scaling like @xmath195 and such that ( ) ( ) = _ 0^ dt - 12 _ 0 ^ 2 _ 0^ dt d. one can check that one has indeed @xmath196 .
furthermore , it follows from their explicit expressions that both @xmath188 and @xmath197 map @xmath198 functions into @xmath198 functions as long as @xmath199 $ ] .
we conclude this section with a small lemma that allows us to compute the @xmath54 norm of functions that scale in a certain way .
let @xmath200 and @xmath201 for some @xmath202 .
the functions that will be considered in the remainder of this section will always be of the form _ ( p , q , x , y ) = f(x - g(p , q ) , y - h(p , q ) ) ( p , q ) ( ( p , q)/ ) , for some parameter @xmath203 .
one has [ lem : norm ] let @xmath204 , let @xmath205 be measurable , let @xmath206 be continuous and compactly supported away from @xmath34 , and let @xmath207 be a function that is continuous away from @xmath34 and scales like @xmath171 for some @xmath208 .
then , the @xmath54-norm of @xmath209 defined as in satisfies @xmath210 .
make the change of variables @xmath211 and use the scaling properties of @xmath207 .
before we tackling the general case of a chain with arbitrary length , let us `` cut our teeth '' on the problem with three oscillators . since we do not have an explicit expression for the invariant measure @xmath15 ( indeed , at this stage , we do not even know that it exists ! ) , we are going to study the generator of in spaces of the type @xmath212 . as in @xcite
, it is not expected that the qualitative nature of the spectrum of @xmath18 depends on the choice of @xmath213 , as long as @xmath214 ( as usual , we set @xmath215 ) .
since one expects the true invariant measure to be somehow in between the gibbs measures at temperatures @xmath27 and @xmath216 , it is very likely that the qualitative nature of the spectrum of @xmath18 in @xmath20 is also the same .
we write h_0 & = p_0 ^ 2 2 + |q_0|^2k 2k + q_0 ^ 2 2 , h_2 = p_2 ^ 2 2 + |q_2|^2k 2k + q_2 ^ 2 2 , + h_1 & = p_1 ^ 2 2 + |q_1|^2k 2k .
one should not think of @xmath217 ( and @xmath218 ) as the energy one gets by pinning @xmath219 at @xmath34 , but rather as the energy such that the corresponding force is the one that gets averaged out over the fast motion of the middle oscillator . with this notation
, the liouville operator @xmath220 for the total hamiltonian can be broken up as follows : x_h = x_h_0 + x_h_2 + x_h_1 + ( q_0 + q_2 - 2q_1)_p_1 + q_1 ( _ p_0 + _ p_2 ) .
recall that the generator of the stochastic dynamics is given by = x_h - _ 0 p_0_p_0 + _ 0 t_0 _ p_0 ^ 2 - _ 2 p_2_p_2 + _ 2 t_2 _ p_2 ^ 2 .
the space @xmath212 isometric to @xmath54 via the operator @xmath221 .
this shows that @xmath18 is conjugate to the operator @xmath222 on the flat @xmath54 space given by & = x_h + _
i=0,2_i t_i ( ( _ i - _ i^*)p_i _ p_i + _ p_i^2 - _ i _ i^ * p_i^2 + _ i ) + & = x_h + _ 0 t_0 _ ^0 + _ 2 t_2 _ ^2 . where we set _ i = 2,_i^ * = 1t_i -2 . here
we see the importance of the condition @xmath223 : it makes sure that the coefficients in front of @xmath224 are strictly negative .
if this is not the case , @xmath225 is not dissipative anymore and does therefore not generate a @xmath226-semigroup on @xmath227 .
the main result of this section is : [ theo : threeosc ] if @xmath228 , then the operator @xmath225 does not have compact resolvent for any @xmath229 .
if @xmath4 , then it has essential spectrum at @xmath34 . the aim is to construct a sequence @xmath155 of _ approximate eigenfunctions _ such that all the @xmath155 are mutually orthogonal , @xmath162 and @xmath230 either stays bounded or converges to @xmath34 . by propositions [ prop : compact ] and [ prop : compact2 ] ,
this would then immediately imply a lack of compactness for the resolvent of @xmath231 , or even the presence of essential spectrum at @xmath34 .
since the spectral properties of @xmath225 and of its adjoint @xmath232 are the same , we can also construct such a sequence of approximate eigenfunctions for @xmath233 instead .
they can then be interpreted as _ approximate invariant measures _ for the dynamic .
since it seems to be a little bit easier to get an intuition about densities of approximate invariant measures rather than about approximately invariant observables , this is what we are going to do in this section .
the adjoint of @xmath225 is given by ^ * & = -x_h + _ 0 t_0 ( _ ^0)^ * + _ 2 t_2 ( _ ^2)^ * + ( _ ^i)^ * & = ( _ i^ * - _ i ) p_i
_ p_i + _ p_i^2 - _ i _ i^ * p_i^2 + _
i^*. direct calculation shows that @xmath234 is an eigenfunction with eigenvalue @xmath34 for @xmath235 and @xmath236 is an eigenfunction with eigenvalue @xmath34 for @xmath237 with @xmath238 .
however , if @xmath239 ( and therefore @xmath240 ) it is not possible in general to find a closed expression for an eigenfunction with eigenvalue @xmath34 for @xmath233 .
note also that @xmath241 if and only if @xmath242 , which is not surprising since in this case @xmath235 is self - adjoint .
choose now a function @xmath243 $ ] which is smooth and compactly supported on @xmath244 $ ] and set for example @xmath245 , so that the functions @xmath246 have disjoint support .
the formal calculation performed in section [ sec : formalcalc ] suggests that when the energy of the middle oscillator is large , the dynamic of keeps that energy approximately constant , while the two boundary oscillators equilibrate approximately at gibbs measures at temperatures @xmath27 and @xmath216 respectively .
our first guess would be therefore to build approximate eigenfunctions for @xmath225 and @xmath233 by setting _
n = c_n e^- _ 0 h_0 - _ 2 h_2 ( h_1/_n),_n^ * = c_n^ * e^- _ 0^ * h_0 - _ 2^ * h_2 ( h_1/_n ) . here , the constants @xmath247 and @xmath248 are chosen such that @xmath249 . from , one infers that @xmath250 . with this guess ,
we get _ n = c_ne^-_0 h_0 - _ 2 h_2 ( p_1 ( q_0 + q_2 - 2q_1)_n (h_1 _ n ) - q_1 ( _ 0 p_0 + _ 2 p_2 ) ( h_1 _ n ) ) , and similarly for @xmath251 . since @xmath252 ( on the support of @xmath155 )
, the first term goes to @xmath34 in @xmath54 .
the second term however goes to @xmath253 because of the factor @xmath219 , so we have to be a little bit more careful in our analysis .
the problem is that we have not so far exploited the fact that we also know approximately what the fast oscillations superimposed over the _ slow _ dynamic of the boundary oscillators look like , see .
these oscillations can be expressed as a function @xmath254 of the middle oscillator , solution to the _ poisson equation _ x _ ( p , q ) = -q . by
, this equation has a unique solution that averages to @xmath34 along orbits of the solutions corresponding to @xmath166 .
furthermore , @xmath254 scales like @xmath255 . in particular , we note that @xmath254 is bounded when @xmath112 and converges to @xmath34 at large energies when @xmath4 .
our next guess is therefore to compensate for these fast oscillations by setting @xmath256 and taking _
n = c_n ( - _ 0 h_0(|p_0 , q_0 ) - _ 2 h_2(|p_2 , q_2 ) ) ( h_1 _ n ) .
observe at this stage that @xmath257 , so that we can make use of when computing @xmath258 .
one has , for @xmath259 : x_h_i _ n & = -_i ( q_i|q_i|^2k-2 + q_i ) _ n , + x_h_1 _ n & = _ 0 |p_0 q_1 _ n + _ 2 |p_2 q_1 _ n , [ e : xh1phi ] + q_1 ( _ p_0 + _ p_2 ) _ n & = - _ 0 |p_0 q_1 _ n - _ 2 |p_2q_1 _ n,[e : crossphi ] + ( q_0 + q_2 - 2q_1 ) _ p_1 _
n & = ( q_0 + q_2 - 2q_1)(-(_0 |p_0 + _ 2 |p_2)_p + p_1 _
n )_n , + _ ^i _ n & = _ i ( ( _ i + _ i^ * ) p_i + _
i ) _ n. here , we have omitted the argument @xmath260 of @xmath254 and the argument @xmath261 of @xmath262 and @xmath263 for the sake of simplicity .
note now that and cancel each other out exactly .
it follows from that x_h_i _ n & _ n^1k - 12 , + ( q_0 + q_2 - 2q_1 ) _
n & ( 1 + _ n^1 2k)(_n^1k - 1 + _
n^-12 ) _ n^12k - 12 , + _
n & _ n^1k - 12 + _ n^2k - 1 _ n^1k - 12 .
note that the exponent @xmath264 appearing in is precisely canceled by the normalization constant @xmath247 .
we also used here the symbol @xmath265 for two expressions @xmath266 as a shorthand for `` there exists a constant @xmath267 such that @xmath268 . ''
it follows from the above bounds that @xmath269 .
it is possible to construct approximate eigenfunctions @xmath156 for @xmath270 similarly by setting _ n^
* = c_n ( - _ 0^ * h_0(|p_0 , q_0 ) - _ 2^ * h_2(|p_2 , q_2 ) ) ( h_1_n ) .
note now that the only difference between @xmath18 and @xmath270 is that one changes the sign of @xmath271 and switches @xmath272 and @xmath273 .
this shows that the cancellation between and still takes place when applying @xmath270 to @xmath156 , so that @xmath274 as above .
if @xmath114 , it follows that there exists a constant @xmath267 such that @xmath275 for every @xmath100 . if @xmath4 , all the exponents appearing the the above expressions are negative , so that @xmath276 . applying propositions [ prop : compact ] and [ prop : compact2 ] concludes the proof of the theorem .
it is clear from the proof that the exact same result also holds for a chain consisting of @xmath277 oscillators instead of @xmath58
. one can construct approximate invariant measures in exactly the same way , but one has the additional freedom of choosing to take the energy of either of the two middle oscillators to be large . in this section
, we consider a chain of length @xmath24 for @xmath278
. we will show that if @xmath279 , then the generator of the dynamic has essential spectrum at @xmath34 . to this end , define similarly as before h_0 & = p_0 ^ 2 + p_1 ^ 2 2 + |q_0|^2k + |q_1|^2k 2k + ( q_0-q_1)^2 + q_1 ^ 2 2 , + h_r & = q_3 ^ 2 2 + _ i=3^n(p_i^2 2 + |q_i|^2k2k ) + _ i=4^n ( q_i - q_i-1)^2 2 , + h_c & = p_2 ^ 2 2 + |q_2|^2k 2 , so that x_h = x_h_0 + x_h_r + x_h_c + ( q_1 + q_3 - 2q_2)_p_2 + q_2 ( _ p_1 + _ p_3 ) .
as in the previous section , we consider the operator @xmath225 on the flat @xmath54 space given by = x_h + _ 0 t_0 _ ^0 + _ n t_n _ ^r . we have [ theo : compactchain ] if @xmath280 and @xmath279 , then the operator @xmath225 has essential spectrum at @xmath34 .
if @xmath281 , it does not have compact resolvent .
as previously , these statements are independent of the value of @xmath282 .
let @xmath254 be defined as in the previous subsection . in this section , we do not only add a corrector term to @xmath283 and @xmath284 , but also to @xmath219 and @xmath285 .
we define @xmath286 to be the solution to the poisson equation @xmath287 and we define new variables @xmath288 and @xmath289 by for @xmath290 and @xmath291 otherwise . with this notation , we set as before _ n & = c_n ( - _ 0 h_0(|p , |q ) - _ n h_r(|p , |q ) ) ( h_c/_n ) . in order to compute @xmath292 , let us first apply @xmath271 to @xmath293 and @xmath294 : x_h |q_0 & = |p_0 , + x_h |p_0 & = - |q_0 ||q_0|^2k-2 + |q_1 - |q_0 - ^(2 ) , + x_h |q_1 & = |p_1 + ( q_0 + q_2 - 2 q_1 ) _
p^(2 ) , + x_h |p_1&= - |q_1||q_1|^2k-2 + |q_0 - 2|q_1 + & + ( |q_1||q_1|^2k-2 - q_1 |q_1|^2k-2 ) + ( q_0 + q_2 - 2q_1 ) _ p + 2^(2 ) .
hence x_h h_0(|p , |q ) & = - |p_0 ^(2 ) - ( |q_1 ||q_1|^2k-1 + 2|q_1 - |q_0 ) ( q_0 + q_2 - 2q_1 ) _ p ^(2 ) [ e : xhh0 ] + & + |p_1 ( |q_1||q_1|^2k-2 - q_1 |q_1|^2k-2 ) + |p_1 ( q_0 + q_2 - 2q_1 ) _
p + 2 |p_1 ^(2 ) + & = r_1 + r_2 + r_3 + r_4 + r_5 , and similarly for @xmath295 by symmetry . we have furthermore x_h h_c = p_2 ( q_1 + q_3 - 2q_2 ) , so that x_h ( h_c / _ n ) = q_1 + q_3 - 2q_2 _
n p_2 (h_c / _ n ) . since @xmath296 , we thus have _ n = ( - _ 0 x_h h_0(|p , |q ) - _
n x_h h_r(|p , |q ) + x_h ( h_c / _ n ) ( h_c/_n ) ) _ n .
we bound the terms appearing in this expression in the same way as in the previous subsection . since @xmath254 scales like @xmath297 and @xmath286 scales like @xmath298 , shows that r_1 _ n & _ n^32k-1 , & r_2 _ n & ( 1 + _
n^12k ) _ n^32k-3 2 _ n^2k - 3 2 , + r_3 _ n & _ n^32k-1 , & r_4 _ n & ( 1 + _
n^12k ) _ n^1k-1 _ n^32k - 1 , + r_5 _ n & _ n^32k-1 , & _ n & ( 1 + _
n^12k ) _ n^-12 _ n^12k - 12 . collecting all these bounds ,
we obtain @xmath299 .
as before , if we set _ n^ * & = c_n ( - _ 0^ * h_0(|p , |q ) - _ n^ * h_r(|p , |q ) ) ( h_c/_n ) , we obtain the same bounds for @xmath300 .
the exponents appearing in all of these bounds are strictly negative whenever @xmath301 , thus concluding the proof of theorem [ theo : compactchain ] .
from now on , we study the case of three oscillators in detail . in this section , we derive an effective dynamic for the outer oscillators that is valid in the regime where most of the energy is located in the center oscillator .
more precisely , we show that there exists a change of variable @xmath302 for @xmath259 such that the equations of motion for @xmath303 decouple ( to leading order ) from the rest of the system , provided that the energy of the middle oscillator is large . as before , we will use throughout this section the symbol @xmath304 for two expressions @xmath266 as a shorthand for `` there exists a constant @xmath267 such that @xmath268 . ''
the constant @xmath267 depends in general on the parameters of the model , but is of course independent of the arguments of the @xmath266 .
[ theo : decouple ] assume that @xmath279 .
there exist smooth functions @xmath305 and @xmath306 depending on @xmath307 such that if we make the change of variables @xmath308 and @xmath309 ( for @xmath259 ) , the equations of motion for @xmath303 are given by d |q_i & = |p_i dt + r_q^i dt + _
q^i dw_i + d |p_i & = - |q_i ||q_i|^2k-2 dt - |q_i dt - _ i |p_i dt + r_p^i dt + _
p^i dw_i for some adapted processes @xmath310 , @xmath311 , @xmath312 and @xmath313 . furthermore , the _ error terms _ @xmath314 and @xmath315 satisfy the bounds for some @xmath316 . here
, the energies @xmath317 are given by @xmath318 .
this theorem gives a change of variables where the high - speed oscillations due to the presence of a breather located on the middle oscillator ( that is the case where @xmath319 ) have been decoupled from the remaining degrees of freedom , leaving an effective averaged out dynamic .
recalling the formal calculation performed in section [ sec : formalcalc ] , we see that when the energy @xmath68 is predominantly concentrated in the central oscillator , then the amplitudes of the oscillations for @xmath101 and @xmath98 , @xmath320 scale to leading order like @xmath321 and @xmath322 respectively .
this indicates that there are natural breakpoints at @xmath114 and @xmath323 .
when @xmath228 , the oscillations of both the @xmath101 and the @xmath98 are bounded as @xmath68 increases , so that they can be removed by a change of variables which is a bounded perturbation of the identity .
when @xmath324 , the amplitude of the oscillations of the @xmath101 increases with @xmath68 , but as long as @xmath325 , the amplitude of the @xmath98 does not .
this growth will cause extra difficulties . if we consider @xmath326 , both amplitudes would grow with @xmath68 , leading to further complications . since our goal is to outline the ideas without seeking the greatest generality , we resist the temptation to analyze all cases and restrict ourselves to the case @xmath301 .
before we proceed , let us compute the expressions @xmath314 and @xmath315 for a generic choice of @xmath305 and @xmath306 .
applying its formula to @xmath293 and @xmath294 , we obtain r_p^i & = v(q_i + _ q^i ) - v(q_i ) + _ q^i + _ i _
p^i + _ p^i + q_1 + r_q^i & = _ q^i - _
p^i = _ i ( 1 + _ p_i _ p^i),_q^i
= _ i _ p_i _ q^i . here and in the sequel we write @xmath327 as a shorthand for @xmath88 and @xmath328 as a shorthand for @xmath329 . _
the case @xmath330 . _
the only `` bad '' term in the equations of motion for @xmath331 is the @xmath219 in the right hand side of the equation for @xmath101 .
the case @xmath228 is much easier than the case @xmath332 since the system is more `` rigid '' in the later case .
one can then simply take @xmath333 and @xmath334 , where @xmath254 is the centered solution to the poisson equation x _ = ( p , q ) - q and @xmath335 , where @xmath336 $ ] is a smooth function such that @xmath337 for @xmath338 and @xmath339 for @xmath340 . making this choice of @xmath305 and @xmath306 in yields _ q^i & = 0 , & r_q^i & = -(p_1 ,
q_1 ) , + _ p^i & = _ i , & r_p^i & = _ i ( p_1 , q_1 ) + _ p ( p_1 , q_1 ) ( q_0 + q_2 - 2 q_1 ) + ( p_1,q_1 ) since the function @xmath254 scales like @xmath341 outside of a compact set , it can be checked easily that the bounds hold , provided that @xmath228 .
note that in this case , @xmath342 and @xmath317 are equivalent in the sense that @xmath343 since @xmath254 is bounded .
therefore all occurrences of @xmath344 in the bounds can be replaced by @xmath345 at the cost of multiplicative constants . _
the case @xmath346 .
_ we are now going to assume that @xmath324 , which the more delicate case .
note that the second and third terms in @xmath310 above satisfy the bounds ( with @xmath344 instead of @xmath345 , but this problem will be dealt with later ) provided that @xmath347 scales like @xmath348 for some @xmath349 .
this is the case when @xmath350 , which is one of the reasons why we restrict ourselves to this case .
therefore , only the terms involving @xmath254 scale worse then the desired bounds on the error terms and need to be eliminated .
this motivates the introduction of the solution @xmath286 to the poisson equation x _
^(2 ) = .
note that @xmath286 scales like @xmath351 outside of a compact set , so that it is bounded if @xmath352 .
it would be tempting at this point to simply subtract @xmath353 to @xmath101 and add @xmath354 to @xmath98 .
this would however introduce correction terms that grow faster than the bounds in .
the trick is to realize that these correction terms are multiplied with terms that go to @xmath34 as the energy of the middle oscillator becomes large .
we therefore multiply @xmath286 with a cutoff function that makes sure that this second round of correction is applied only when the energy of the middle oscillator is large compared to the energy of the boundary oscillators .
let @xmath355 $ ] be a smooth increasing function such that @xmath356 for @xmath357 and @xmath358 for @xmath359 .
let furthermore @xmath180 be an exponent to be determined later and set _ p^i & = ( p_1 , q_1 ) - _ i ^(2)(p_1 , q_1 ) , [ e : phip ] + _ q^i & = ( e_1 / e_i^ ) ^(2)(p_1 , q_1 ) , [ e : phiq ] where we defined @xmath360 . in the sequel
, we are going to use the shorthand @xmath361 and we will omit the arguments of @xmath254 , @xmath286 and @xmath362 in order to simplify notations . before we turn to the verification of the bounds , we remark that since @xmath324 , @xmath342 and @xmath317 are _ not _ equivalent for @xmath259 .
since @xmath325 , @xmath363 and @xmath364 are however equivalent in the sense that @xmath365 .
since we wish to bound the remainder in terms of powers of @xmath317 and not @xmath342 , we are now going to show how these quantities are related . from the definitions of @xmath293 and @xmath294
, we have for @xmath320 the estimate by choosing @xmath366 sufficiently small and moving the term @xmath367 to the left hand side , we thus obtain the two bounds e_i |e_i + e_1 ^ 2k - 1,|e_i e_i + e_1 ^ 2k - 1 .
writing @xmath368 , and applying it s formula to @xmath369 , we obtain dg & = e_i^- p_1 ( q_0 + q_2 - 2q_1 ) dt - e_1 e_i^-- 1 ( _ i^22- _ i p_i^2 + q_1 - q_i ) dt + & + ( + 1 ) 2 e_1 e_i^-- 2_i^2 p_i^2 dt - e_1 e_i^--1 _
i p_i dw_i + & = i_1 dt + i_2 dt + i_3 dt + dm(t ) .
@xmath370 is a martingale since , on any finite time interval , it is easy to get control over the expected value of any power of the total energy .
( even a bound which grows exponentially with the length of the time interval is sufficient . ) bounding each of the three terms @xmath371 separately we obtain : 2k , + applying it s formula to @xmath362 , we get _ i^&= (e_1/e_i^ )
g + ^2_i^2 ( e_1/e_i^ )
e_1 ^ 2 e_i^-2 - 2 p_i^2 = i_4 + i_5 .
note now that @xmath372 and @xmath373 are zero outside the interval @xmath244 $ ] .
therefore , we have @xmath374 on the support of these functions .
this line of reasoning yields for @xmath375 the bounds : 1 + e_i^-1 e_2-i^1 2k , + collecting these estimates and taking into account the support of @xmath372 and @xmath373 produces @xmath376 .
recall that , from the scalings given in lemma [ lem : scaling ] of the solutions to the poisson equation and the definitions and of @xmath254 and @xmath286 , we have that @xmath377 , @xmath378 , @xmath379 , @xmath380 .
we will also use the fact that , from the definition of @xmath362 combined with , one has both bounds e_i^e_1,|e_i^e_1 , on the support of @xmath362 .
furthermore , one has the bound @xmath381 on the support of @xmath382 which , by , implies that one also has @xmath383 on the support of @xmath382 . using the definitions of @xmath305 and @xmath306 given in , equation yields r^i_q = - + _
i^ ^(2 ) + ( _ i^ ) ^(2 ) + _ i ^(2 ) .
we now make use of the definitions of @xmath18 and @xmath286 to obtain ^(2 ) = + ( q_0 + q_2 - 2q_1 ) _
p ^(2 ) .
this allows us to obtain the following bounds for @xmath384 : ^(2)| , + & ( 1 - _ i^ ) e_1 ^ 1 k - 1 2 + _
i^e_1 ^ -32 ( 1 - 1 k ) ( e_i^1 2k+e_1 ^ 1 2k+e_2-i^1 2k ) + & + e_2-i^1 2k e_1 ^ 3 2k - 1_2e_i^e_1 e_i^ + 1 + & = i_6 + i_7 + i_8 + 1 .
our aim is to bound the terms @xmath385 , @xmath386 , @xmath387 in terms of the @xmath345 s instead of the @xmath344 s .
to do so , we now fix = 32 . since @xmath388 , @xmath389 which is the constraint that we had to impose earlier .
since , as mentioned above , @xmath390 on the support of @xmath382 , one has one can check that the choice @xmath391 implies that there exists @xmath316 so that @xmath392 ( actually , one can set @xmath393 for the range of values of @xmath86 considered here ) .
this shows that @xmath394 as required . to bound @xmath386 , we make use of the fact that e_i^1 2k + e_2-i^1 2k |e_i^1 2k + e_1 ^ 1k^2 - 1 2k + |e_2-i^1 2k |e_i^1 2k + e_1 ^
1 2k + | e_2-i^1 2k , so that , by virtue of , & |e_i^- ( |e_i^1 2k+ | e_2-i^1 2k ) + 1 . we finally turn to @xmath387 .
we first remark that , by and the fact that @xmath395 , one has @xmath396 on the support of the indicator function @xmath397 where @xmath398 .
we thus obtain - _ a + & ( |e_2-i^1 2k + |e_i^ 2k(2k - 1 ) ) |e_i^32k - |e_2-i^1 2k |e_i^- + | e_i^ .
one can check that , for @xmath391 and the range of @xmath86 s of interest , the exponent of the last term is strictly smaller than @xmath399 .
therefore , @xmath400 as required , choosing @xmath401 smaller if necessary . collecting all of these bounds
shows that @xmath311 does indeed satisfy the bound advertised in .
turning to @xmath310 , yields & + v(q_i + _
i^ ^(2 ) ) -v(q_i)| , + & 1 + ( |e_0 ^ 1 2k + |e_2 ^ 1 2k + e_1 ^ 1 2k)e_1 ^ 1 k-1 + 1 + |v(q_i + _
i^ ^(2 ) ) -v(q_i)| . here , we made use of and of .
since , for @xmath325 , @xmath286 is a bounded function , one has the further bound & ( 1+||q_i|)^2k-2 |^(2)| |e_i^1 - 1
k |e_i^1 2 - , for @xmath401 sufficiently small .
collecting these bounds and using the fact that @xmath301 , we obtain which is indeed of the form .
lastly , the @xmath315-terms can be bounded by _ p^i & = _ i , where we made use of the fact that , like in the bound of @xmath387 , @xmath363 , @xmath402 and @xmath403 are equivalent on the support of @xmath404 .
this concludes the proof of theorem [ theo : decouple ] . in the case
@xmath405 , it will be useful in the sequel to have a better approximation of the dynamics that yields smaller error terms in the regime where most of the energy is located in the middle oscillator : [ theo : effective2 ] assume that @xmath4 .
there exist smooth functions @xmath305 and @xmath306 depending on @xmath307 such that if we make the change of variables @xmath308 and @xmath406 ( for @xmath259 ) , the equations of motion for @xmath407 are given by d |q_i & = |p_i dt + r_q^i dt + d |p_i & = - |q_i ||q_i|^2k-2 dt - |q_i dt - _ i |p_i dt + r_p^i dt + _
i dw_i for some adapted processes @xmath310 and @xmath311 .
furthermore , the _ error terms _
@xmath314 satisfy the bounds here , the energies @xmath342 are given as before by @xmath408 and @xmath38 is the total hamiltonian of our system .
following the proof of theorem [ theo : decouple ] , we set as in _ p^i & = ( p_1 , q_1 ) - _ i ^(2)(p_1 , q_1 ) , + _ q^i & = ^(2)(p_1 , q_1 ) .
this yields for @xmath310 and @xmath311 the expressions r^i_q & = ( q_0 + q_2 - 2q_1 ) _
p ^(2 ) + _ i ^(2 ) , + r_p^i & = ( q_1 ) + ( q_0 + q_2 - 2q_1 ) ( _ p- _ i _
p^(2 ) ) - ( _ i^2 - 1 ) ^(2 ) + & + v(q_i + ^(2 ) ) -v(q_i ) .
the desired bounds now follow from lemma [ lem : trivial ] below , together with the fact that both @xmath286 and @xmath347 scale like @xmath409 .
[ lem : trivial ] for every @xmath410 , the bound x^- ( x+y)^- \{2^ , y^ } holds for every @xmath411 .
if @xmath412 , then @xmath413 . if on the other hand @xmath414 , then @xmath415 .
to prove the existence of an invariant measure , our aim is to construct a type of _ lyapunov function _
@xmath416 such that @xmath417 as @xmath418 and @xmath419 as @xmath418 .
given such a function , the existence of an invariant measure will follow from the following proposition which is a variant of the classical kryloff - bogoliouboff construction @xcite : [ abstractexistinvmeasure ] consider an sde on @xmath420 with smooth coefficients and denote its generator by @xmath18 .
assume that the sde has global solutions and generates a feller semigroup .
if there exists a smooth function @xmath421 such that the level sets @xmath422 are compact for every @xmath267 , then the sde possesses an invariant probability measure @xmath15 .
furthermore , the function @xmath423 is integrable with respect to @xmath15 and @xmath424 .
we will construct the function @xmath425 in steps by analyzing the dynamic in the limit of various energies being large and then draw inspiration from the structure of these limiting regimes to construct @xmath425 .
operationally , we will make an initial guess for @xmath425 and then augment it by a series of correction terms .
now , by assumption the sets @xmath436 are compact for all @xmath314 . in particular , this implies that there exist a @xmath437 so that @xmath438 for all @xmath125 .
now observe that for @xmath439 @xmath440 therefore , the sequence of measures @xmath441 defined for measurable sets @xmath442 by @xmath443 is tight .
hence the kryloff - bogoliouboff construction @xcite guarantees the existence of an invariant measure .
the last statement then follows from lebesgue s dominated convergence theorem . for some positive constants @xmath444 and some ( typically quite small ) @xmath445 $ ] . in this case , it is known @xcite that one does not only have the existence of an invariant measure , but the transition probabilities converge towards it at rate @xmath446 in the total variation distance .
we believe that this convergence actually takes place at a much faster rate , but such a statement is beyond our reach at the moment .
see also @xcite for related results on subexponential mixing for sdes .
recall the change of variables @xmath447 from theorem [ theo : decouple ] that leads to an effective decoupled dynamic for the outside oscillators . in order to construct the lyapunov function @xmath425
, we proceed in two steps : for some constant @xmath111 to be determined later .
if @xmath111 is sufficiently small and since @xmath454 , this function is indeed equivalent to @xmath455 . applying it s formula to it , we get from that [ theo : wholething ] consider the equations of motion with @xmath301
. then there exists a function @xmath425 and constants @xmath462 such that @xmath463 and such that @xmath464 for some exponents @xmath180 and @xmath465 .
if @xmath473 , then the motion of the interior oscillator induces large ( going to infinity with the energy of the middle oscillator ) fluctuations in the values of @xmath474 and @xmath475 .
therefore , there will always be energy dissipation . if @xmath228 , then the motion of the interior oscillator induces bounded ( or small ) fluctuations in the values of @xmath474 and @xmath475 . in this case , one can subtract a compensator so that only fluctuations remain .
there will be a few error terms that seem to be larger than the dominant one , but they can hopefully just be eliminated order by order . _
the case @xmath473 : _ in contrast to the arguments in the proof of theorem [ theo : decouple ] , this is the `` easy '' case for this part of the proof .
it follows from the proof of theorem [ theo : decouple ] that in this case @xmath476 is equivalent to @xmath477 with @xmath478 .
since @xmath479 , this allows us to obtain for @xmath480 the bound note that @xmath481 scales like @xmath482 which goes to infinity at high energies .
we can thus find a positive constant @xmath84 ( which was introduced in ) and a function @xmath483 such that @xmath484 averages out to @xmath34 in the sense of proposition [ prop : scale ] .
we define @xmath485 to be the solution to the poisson equation this function is smooth since @xmath483 is precisely such that @xmath486 in a neighborhood of the origin .
the function @xmath485 then scales like @xmath487 . in particular , @xmath485 and @xmath488 are bounded .
we can check that @xmath466 satisfies the first bound in , since @xmath485 grows slower than @xmath38 and @xmath448 is bounded from below by a positive constant .
however , it is not clear _ a priori _ from this expression that it @xmath466 satisfies the second bound in . _
the case @xmath494 : _ the reasoning above only worked for @xmath324 since in this case one has @xmath495 . when @xmath494 the situation is slightly more delicate . here ,
as before , the energy dissipation mechanism comes from fluctuations of @xmath496 and @xmath497 around their mean .
however , the amplitude of these oscillations now _ decreases _ as the energy stored in the middle oscillator increases , so that we have to be more careful .
in particular , @xmath224 can no longer be treated as a perturbation with respect to @xmath481 , as we did in the previous case . where the `` bar '' variables are as in theorem [ theo : effective2 ] .
note that here , and anywhere from this point on , we use the variables @xmath303 from theorem [ theo : effective2 ] and _ not _ from theorem [ theo : decouple ] .
this is because we require a sufficiently good effective dynamics so that the error terms are small with respect to @xmath481 .
the bounds obtained in theorem [ theo : decouple ] are not sufficiently small for that . setting as before @xmath317 as a shorthand for @xmath498
, one has this puts us now in a situation similar to , with the difference that , up to powers of @xmath448 , the error term @xmath503 is of order @xmath505 instead of being of order @xmath506 .
since @xmath481 is larger than @xmath507 , this is the feature that will allow us to obtain the required bound .
since these terms oscillate very rapidly , we would like to replace them by their averaged effect over one period of the middle oscillator .
to leading order , the terms @xmath511 and @xmath481 will average out to @xmath34 and @xmath512 respectively .
the latter contribution will turn out to be the dominant term leading to an overall dissipation of energy . defining @xmath485 and @xmath286 as in and
respectively , we finally set h^n-1 |p_i ^(2 ) & = h^n-1|p_i + h^n-1|p_i ( q_0 + q_2 - 2q_1)_p ^(2 ) + & + ( n-1)h^n-3 |p_i ^(2)_j=0,2 ( h(_j^2 2 - _ j
p_j^2 ) + ( n-2)_j^2 p_j^2 ) + & + h^n-1^(2 ) ( - |q_i||q_i|^2k-2 - | q_i + r_p^i - _
i |p_i ) + ( n-1)h^n-2_i^2 p_i ^(2 ) + & = h^n-1|p_i + i_11 + i_12 + i_13 + i_14 .
eckmann , c .- a .
pillet , and l. rey - bellet .
non - equilibrium statistical mechanics of anharmonic chains coupled to two heat baths at different temperatures .
phys . _ * 201 * , no . 3 , ( 1999 ) , 657697 . r. z. hasminski . _ stochastic stability of differential equations _ , vol
. 7 of _ monographs and textbooks on mechanics of solids and fluids : mechanics and analysis_. sijthoff & noordhoff , alphen aan den rijn , 1980 .
translated from the russian by d. louvish .
e. hairer , c. lubich , and g. wanner .
_ geometric numerical integration _ , vol .
31 of _ springer series in computational mathematics_. springer - verlag , berlin , second ed . , 2006 .
structure - preserving algorithms for ordinary differential equations . | 1em we study the dynamic behavior at high energies of a chain of anharmonic oscillators coupled at its ends to heat baths at possibly different temperatures . in our setup , each oscillator is subject to a homogeneous anharmonic pinning potential @xmath0 and harmonic coupling potentials @xmath1 between itself and its nearest neighbors .
we consider the case @xmath2 when the pinning potential is stronger then the coupling potential . at high energy ,
when a large fraction of the energy is located in the bulk of the chain , breathers appear and block the transport of energy through the system , thus slowing its convergence to equilibrium .
in such a regime , we obtain equations for an effective dynamics by averaging out the fast oscillation of the breather . using this representation and related ideas
, we can prove a number of results .
when the chain is of length three and @xmath3 we show that there exists a unique invariant measure . if @xmath4 we further show that the system does not relax exponentially fast to this equilibrium by demonstrating that zero is in the essential spectrum of the generator of the dynamics .
when the chain has five or more oscillators and @xmath3 we show that the generator again has zero in its essential spectrum .
in addition to these rigorous results , a theory is given for the rate of decrease of the energy when it is concentrated in one of the oscillators without dissipation .
numerical simulations are included which confirm the theory .
= 0.65 cm |
the quantum electrodynamical model involving the one - photon interaction of a single quantized mode of the electromagnetic field with a motionless two - level atom , known popularly as the jaynes - cummings ( jc ) model @xcite in the literature , has been extensively studied in the last four decades .
the jc model has predicted many unexpected and non - trivial results , e.g. the well known phenomenon of collapses and revivals of rabi oscillations- unquestionable evidence of the quantum nature of the electromagnetic field @xcite .
the other non - trivial and fascinating quantum features displayed by the jc model are squeezing of the cavity - field @xcite , atomic dipole squeezing @xcite and atom - cavity entanglement @xcite . at present
, the jc model occupies a special place in quantum optics because it helps to examine and verify the conjectures about more complicated models which are close to real processes .
recent experiments with rydberg atoms and microwave photons in a superconducting cavity have turned the jc model from a theoretical curiosity to a useful and testable enterprise @xcite .
such a system is also suitable for quantum - state engineering and quantum information processing . up to now ,
fock states @xcite , schrdinger cat states @xcite , and entangled states @xcite have been produced in cavity quantum electrodynamics experiments .
stimulated by the success of the ( standard ) jc model , it has been generalized and extended in many interesting directions to explore new quantum effects .
discussions related to several interesting generalizations of this model are now available in the literature @xcite . among the generalized versions of the jc model the
so - called _ f_-deformed jc model ( see , e.g. refs .
@xcite , and @xcite ) has received much attention in view of its connection with quantum algebras @xcite .
the quantum algebras , which appear as a generalization of the symmetry concept @xcite and the basics of so - called non - commutative theories , have given the possibility of generalizing the notion of creation and annihilation operators of the usual quantum oscillator and to introduce deformed oscillator .
some deformed versions of oscillator algebra have found many applications to various physical problems , such as the treatment of vibrational spectra of molecules @xcite and the investigation of nonlinearities in quantum optics @xcite .
in addition , it has been shown @xcite that most of the nonlinear generalizations of the jc model are only particular cases of the _ f_-deformed jc model in which the creation and annihilation operators of the radiation field are replaced by deformed harmonic - oscillator operators with prescribed commutation relations .
the representation theory of the quantum algebras with a single deformation parameter _
q _ has led to the development of the _ q_-deformed oscillator algebra @xcite . by using a _
q_-oscillator description , the jc model hamiltonian with an intensity - dependent coupling has been generalized by relating it to the quantum @xmath0 algebra @xcite .
it has been found @xcite that the periodicity of the time evolution of the physical quantities , which is a typical feature of the jc model with intensity - dependent coupling , is destroyed when the atom interacts with the @xmath1 analogue of the @xmath2 coherent state . in order to extract possible information about the physical meaning of the _ q_-deformation , the atomic inversion of the standard jcm with a _
q_-deformed field initially prepared in a _
q_-deformed coherent state @xcite has been studied @xcite .
various versions of the jc model and their _ _ q-__deformed extensions have been treated in a unified formalism based on the generalized deformed oscillator algebra @xcite .
the temporal evolution of atomic inversion and quantum fluctuations of atomic dipole variables in three variants of the two - photon non - dissipative _ q_-deformed jc model have been studied @xcite for both on - and off - resonant atom - field interaction .
particularly , it has been found that for nonzero detuning the atomic inversion exhibits superstructures , which are absent in the non - deformed jc model , and the magnitude of dipole squeezing may be increased . by solving a dynamical problem characterized by a quite general _ f_-deformed jc model , it has also been proposed @xcite a theoretical scheme to show the possibility of generating various families of nonlinear ( _ f_-deformed ) coherent states @xcite of the radiation field in a lossless coherently pumped micromaser . in ref .
@xcite , the influence of nonlinear quantum dissipation on the dynamical properties of the _ f_-deformed jc model in the dispersive approximation has been explored . in ref .
@xcite the author has studied the influence of the intrinsic decoherence on quantum statistical properties of a nondegenerate two - photon _ f_-deformed jc model .
furthermore , by introducing some classes of two - parameter ( _ p , q _ ) deformation of the jc model in the rotating wave approximation ( rwa ) the associated nonlinear vector coherent states have been constructed @xcite .
another significant and noteworthy extension of the jc model is the inclusion of atomic motion as well as the effect of the field structure of the cavity mode sustained in the cavity ( see , e.g. refs .
the motivation for such studies is the recent development in the technologies of laser cooling and atom trapping to achieve cooled atoms and super - cooled atoms @xcite . on the other hand , when the atomic motion is taken into account
, some interesting nonlinear transient effects that are similar to self - induced transparency and adiabatic following can be observed in the atomic population @xcite .
furthermore , the influences of the atomic motion and the field - mode structure on the squeezing of the field @xcite , on the cavity - field entropy and schrdinger - cat states generation in the jc model @xcite , and on the rabi oscillations of an atom irradiated by a laser in resonance with the atomic transition @xcite have been studied . in ref .
@xcite the authors have provided a dressed formalism for the dynamics of a moving two - level atom interacting with a quantized cavity - field by taking into account the roentgen interaction @xcite .
all of the foregoing studies devoted to the jc model with atomic motion have been done only under the condition that the influence of the gravitational field is not taken into account .
however , it has often been argued that the achievement of the cooled atoms would be a formidable task as the atoms move so slowly ( with a velocity of a few millimeters or centimeters per second for a time period of several milliseconds or more ) that they start exactly to be extremely sensitive to earth gravity @xcite . for this reason ,
it is of interest to study the temporal evolution of a moving atom simultaneously exposed to the gravitational field and a single - mode cavity - field . by referring to the equivalence principle
, one can get a clear picture of what is going to happen in the interacting atom - field system exposed to a classical homogeneous gravitational field @xcite .
gravity action on particles of small velocities has been observed in matter wave experiments @xcite .
recently , it has been identified @xcite the lowest stationary quantum state of neutrons in the earth s gravitational field by measuring the neutron transmission between a horizontal mirror on the bottom and an absorber or scatterer on top .
a semi - classical description of a two - level atom interacting with a running laser wave in a gravitational field has been studied in refs .
however , the semi - classical treatment prevent us from studying the pure quantum effects occurring in the course of atom - field interaction . in a series of recently published papers @xcite various aspects of the influence of a classical homogeneous gravitational field on quantum dynamics of the jc model
have been analyzed . in ref .
@xcite , within a quantum treatment of the internal and external dynamics of the atom , a theoretical scheme has been presented based on an @xmath3 dynamical algebraic structure to investigate the influence of a classical homogeneous gravitational field on the quantum non - demolition measurement of atomic momentum in the dispersive jc model .
the effects of the gravitational field on quantum statistical properties of the lossless as well as the phase - damped jc models have been studied in refs .
@xcite and @xcite , respectively .
also , the effects of gravity on the dynamical evolution of the cavity - field entropy and the creation of the schrdinger - cat state in the jaynes - cummings model have been examined in ref .
the results reveal that the gravitational field seriously suppresses non - classical properties of both the cavity - field and the moving atom . motivated by the above - mentioned studies on various versions of generalized jc models , in the present paper we deal with the question of how does the gravitational field affect nonclassical statistical properties of the _
f_-deformed(nonlinear ) jc model .
the system under consideration consists of a moving two - level atom exposed to a classical homogeneous gravitational field and interacting with an _
f_-deformed single - mode quantized cavity - field including the nonlinearities of both the field and the intensity - dependent atom - field coupling .
formally , the model hamiltonian for the atom - field system has the same structure as non - deformed jc model in the presence of gravity @xcite with the photon annihilation and creation operators @xmath4 and @xmath5 replaced , respectively , by the _
f_-deformed operators @xmath6 and @xmath7 obeying the deformed commutation relation @xmath8 = ( \hat n + 1)f^2(\hat n + 1 ) - \hat n f^2(\hat n)$ ] .
the nonlinearity ( deformation ) @xmath9 function plays an important role in our treatment since it determines the form of nonlinearities of both the cavity - field and the intensity - dependent atom - field coupling .
this paper is organized as follows . in section 2
, we present a quantum treatment of the internal and external dynamics of the moving two - level atom and find an alternative _
f_-deformed @xmath3 dynamical algebraic structure within the system . based on this deformed @xmath3 structure ,
we obtain an effective hamiltonian describing the nonlinear atom - field interaction in the presence of a classical homogeneous gravitational field . in section 3 , by solving the schrdinger equation in the interaction picture , we investigate the dynamical evolution of the system under consideration and show that how the gravitational field may affect the dynamical properties of the jc model . in section 4 ,
we study the influence of the classical gravitational field on the quantum statistical properties of the two - level atom and the _ f_-deformed quantized radiation field . considering the cavity - field to be initially in a _
q_-deformed coherent state , associated with the nonlinearity function @xmath10 , and the two - level atom in a coherent superposition of the ground and excited states , we investigate the temporal evolution of the atomic population inversion , atomic dipole squeezing , atomic momentum diffusion , photon counting statistics , and deformed quadrature squeezing of the radiation field .
finally , we summarize our conclusions in section 5 .
we consider the interaction of a single two - level moving atom of mass @xmath12 characterized by the pauli operators @xmath13 , the momentum and position operators @xmath14 and @xmath15 with an _
f_-deformed single - mode traveling field characterized by the _
f_-deformed annihilation and creation operators @xmath16 , respectively , in the presence of a classical gravitational field . following refs .
@xcite , the total hamiltonian of the system under the rwa can be written as @xmath17 where @xmath18 denotes the atomic transition frequency , @xmath19 is the frequency of the radiation field , the atom - field coupling constant @xmath20 is a real number , @xmath21 is the wave - vector of the traveling radiation field , and @xmath22 denotes the constant gravitational acceleration acting on the atom .
the operators @xmath16 are deformed annihilation and creation operators constructed from the usual bosonic operators @xmath23 = 1)$ ] and number operator @xmath24 as follows @xmath25 in which the deformation function @xmath26 is an arbitrary real function of @xmath27 .
the deformed operators @xmath16 satisfy the deformed bosonic oscillator commutation relations @xmath28 = ( \hat n + 1)f^2(\hat n + 1 ) - \hat n f^2(\hat n),~ [ \hat a,\hat n ] = \hat a,~ [ \hat a^ + , \hat n ] = - \hat a^ + .
\end{array}\ ] ] the knowledge of the function @xmath26 determines all the properties of the deformed algebra ( 3 ) . in the limiting case @xmath29 , the hamiltonian ( 1 ) reduces to the non - deformed jc model in the presence of atomic motion and gravity @xcite and the algebra ( 3 ) reduces to the well - known heisenberg - weyl algebra generated by @xmath30 and the identity @xmath31 : @xmath32 = \hat i,~ [ \hat a,\hat n ] = \hat a,~ [ \hat a^ + , \hat n ] = - \hat a^ + .
\end{array}\ ] ] to achieve a more clear insight to the physical meaning of the _ f_-deformed jc hamiltonian given in eq.(1 ) it is proper to rewrite it in terms of non - deformed field operators @xmath30 . using ( 2 ) we arrive at @xmath33 in this way
, we learn that the above hamiltonian describes an intensity - dependent interaction of a single two - level moving atom , exposed to a classical homogeneous gravitational field , with a non - deformed single - mode radiation field in the presence of an additional nonlinear interaction represented by the fifth term in eq.(5 ) . as a well - known example if we choose @xmath34 where @xmath35 is a positive constant , the fifth term in eq.(5 ) takes the form @xmath36 which is reminiscent of the kerr - induced interaction @xcite with @xmath37 as the dispersive part of the third - order nonlinearity of the kerr - like medium .
therefore , in this case the model consists of a single two - level moving atom exposed simultaneously to a classical homogeneous gravitational field and a single - mode field surrounded by a nonlinear kerr - like medium contained inside a lossless cavity .
furthermore , it is easily seen that with the above choice for the function @xmath26 the commutation relations ( 3 ) take the form @xmath8 = 2\hat a_0,\ , [ \hat a_0 , \hat a^ + ] = k\hat a^ + , \,[\hat a_0 , \hat a ] = - k\hat a$ ] , with @xmath38 .
it is noteworthy that these relations define @xmath39 algebra when @xmath40 .
it is apparent that there exist a constant of motion in hamiltonian ( 5 ) , @xmath41 that is @xmath42 = [ \hat c,\hat n ] = [ \hat c,\hat\sigma _ z ] = 0 .
\end{array}\ ] ] by using the constant of motion @xmath43 , which represents the total number of excitations of the atom - radiation system , one can expand any function @xmath44 as follows @xmath45 by considering @xmath46 we obtain @xmath47 by making use of eq .
( 9 ) in the hamiltonian ( 5 ) and after some rearrangement , we arrive at @xmath48 where @xmath49 and @xmath50 we now introduce the following operators @xmath51 which satisfy the following commutation relations @xmath52
= - 2\xi ( \hat n;\left| e \right\rangle \left\langle e \right|;\left| g \right\rangle \left\langle g \right|)\hat s_0 , ~[\hat s_0 , \hat s _
^d ] = \pm \hat s _
^d , \end{array}\ ] ] where @xmath53 the operators @xmath54 are the generators of an _
f_-deformed @xmath3 algebra . in the terms of these generators ,
the hamiltonian ( 10 ) can be rewritten as @xmath55 the dynamical evolution of the system can be investigated by the hamiltonian ( 16 ) . for solving eq.(16 )
we introduce the time - dependent unitary operator @xmath56 and we have @xmath57 where @xmath58 and @xmath59 here we have introduced the operator @xmath60 as the _ f_-deformed doppler shift detuning at time _
t _ , @xmath61 thus , due to the presence of the gravitational field ( @xmath62 ) , the nonlinear associated with _ f_-deformation ( @xmath9 ) , and the doppler shift ( @xmath63 ) the detuning between the cavity - field and the atomic transition frequency is modified .
finally , in the interaction picture the transformed hamiltonian ( 20 ) takes the following form @xmath64 where @xmath65 . by using the above hamiltonian we can investigate the dynamical evolution of the system under consideration
this will be done in the two following sections .
in the previous section , we obtained an effective hamiltonian for the interaction of a two - level atom with an _
f_-deformed quantized radiation field , in the presence of a classical homogenous gravitational field . in the present section ,
we investigate the temporal evolution of the system . for this purpose
, we proceed to solve the schrdinger equation for state vector @xmath66 ; i.e. @xmath67 at any time @xmath68 , the state vector @xmath66 is a linear combination of the states @xmath69 and @xmath70 . here
@xmath71(@xmath72 ) is the state in which the atom is in the excited state @xmath73(the ground state @xmath74 ) and the cavity - field has @xmath75 photons .
now we can propose the following ansatz for @xmath66 @xmath76}. \end{array}\ ] ] since the interaction hamiltonian @xmath77 couples only the states @xmath78 and @xmath79 , we consider the evolution of the amplitudes @xmath80 . by substituting equation ( 25 ) into ( 24 ) , the following equations of motion for the time - dependent probability amplitudes @xmath80 are obtained @xmath81 where , we have introduced the modified detuning parameter @xmath82 as @xmath83 with @xmath84 and @xmath85 is the usual detuning parameter . +
the two coupled first - order differential equations ( 26 ) can be solved in a straightforward way , so we have @xmath86 where @xmath87 and @xmath88 denote , the hermit and the confluent hyper - geometric functions , respectively , and by definition @xmath89 in which @xmath90 with @xmath91 , is the _
f_-deformed gravity - dependent rabi frequency . at time
@xmath92 the atom is uncorrelated with the field and the state vector of the system can be written as a direct product @xmath93 where we have assumed that initially the cavity - field is in a coherent superposition of fock states ( @xmath94 ) , the two - level atom is in a coherent superposition of its excited and ground states , and the state vector for the center - of - mass degree of freedom is @xmath95 .
+ thus , the time - independent coefficients @xmath96 and @xmath97 in eqs.(30 ) are obtained as @xmath98 in the next section , by using the sate vector ( 25 ) , we examine the temporal evolution of quantum statistical properties of the _ f_-deformed interacting atom - field system in the presence of a classical homogeneous gravitational field .
in this section we shall study the influence of the classical gravity field on the dynamical evolution of quantum statistical properties of the _ f_-deformed jc model , including the atomic population inversion , atomic dipole squeezing , atomic momentum diffusion , photon - counting statistics , and deformed quadrature squeezing of the field . here
, we discuss the collapse - revival phenomenon in the evolution of the atomic inversion for the state vector @xmath99 .
this phenomenon is a purely quantum mechanical effect and has its origin in the granular structure of the photon number distribution of the initial field @xcite . the phenomenon has been realized experimentally in the sense that the state of the atomic beam leaving the cavity is monitored by ionization detectors @xcite .
the revival - collapse phenomenon has also been seen in nonlinear optics for the single - mode mean photon number of the kerr nonlinear coupler @xcite when the modes are initially prepared in coherent state . at present
this phenomenon is actively discussed in quantum optics not only for atomic system applications but also for the cases of low - dimension quantum nanostructures @xcite , the landau - zener process @xcite , and trapped atoms @xcite . for the model under consideration the atomic population inversion at time @xmath68
is obtained as follows @xmath100 we assume at @xmath92 , the two - level atom is in a coherent superposition of the excited state and the ground state with @xmath101 and the cavity - field is prepared in a
_ maths_-type _
q_-deformed coherent state , @xmath102 in which @xmath103 ! } } } } $ ] with @xmath104 ! } } } } \right)^ { - 1/2}$ ] and @xmath10 .
this type of deformed coherent states can be identified as a family of nonlinear coherent states which can be generated in a lossless coherently pumped micromaser with an intensity - dependent atom - field coupling governed by the nonlinearity function @xmath105 @xcite . from the point of view of the photon statistics
, this family of states exhibits some nonclassical features such as antibunching and quadrature squeezing for certain ranges of the deformation parameter _
+ in figs.1(a)-1(c ) we have plotted the behaviour of @xmath106 as a function of the scaled time @xmath107 for three different values of the parameter @xmath108 and for @xmath109 . in this figure and all the subsequent figures we set @xmath110 and @xmath111 @xcite . here , it is necessary to point out that the relevant time scale introduced by the gravitational influence is @xmath112 @xcite .
thus for an optical field with @xmath113 we have @xmath114 .
it is important to note that the single deformation parameter _
q _ determines the structure of initial deformed coherent state of the cavity - field .
furthermore , it may be viewed as a phenomenological constant controlling the strength of intensity - dependent atom - field coupling as well as the interaction between the field and nonlinear medium contained inside the cavity . as an interesting point
we note that when @xmath115 , the nonlinearity function @xmath9 reduces to @xmath116 .
therefore one can infer that up to the first order approximation the nonlinearity of the model under consideration may be described as a kerr - type nonlinearity .
figure.1(a ) displays the case when the gravitational influence is negligible ( @xmath117 ) .
this means very small @xmath108 , i.e. , the momentum transfer from the radiation field to the atom is only slightly altered by the gravitational acceleration because the latter is very small or nearly perpendicular to the wave - vector @xmath118 .
figures.1(b ) and 1(c ) illustrate the case when we consider the gravitational influence for @xmath119 and @xmath120 , respectively . as it is seen from fig.1(a ) , in the absence of gravitational influence the rabi - like oscillations can be identified .
furthermore , the atomic inversion oscillates around a positive nonzero value which means that due to the deformation more energy is stored in the atomic subsystem .
physically , it is due to the change in energy - level structure of the deformed model under consideration .
increasing of the gravitational influence ( figs.1(b ) , 1(c ) ) induces the population inversion to oscillate so drastically that the phenomenon of collapse - revival is not so clear .
the emergence of this quasi - chaotic behaviour of atomic population inversion can be attributed to a detuning that varies linearly with time [ eq.(27 ) ] and a modified gravity - dependent rabi frequency [ eq.(32 ] .
it has been shown @xcite that these continuous changes in oscillation frequencies influence the collapse and revival times such that subsequent revivals overlap and ultimately a chaotic - like behaviour of time evolution of population inversion occurs .
in addition , with the increasing value of the parameter @xmath108 the atomic inversion oscillates around a negative nonzero value , i.e. , due to the gravitational field more energy is stored in the cavity - field .
besides , numerical analysis shows that by increasing the deformation ( nonlinearity ) parameter _ q _ , the population inversion shows chaotic - like behaviour , for both cases of @xmath117 and @xmath121 .
the origin of this chaotic - like behaviour can also be understood by examining the effect of deformation on the collapse and revival times @xcite .
the squeezing of quantum fluctuations is one of the most fundamental manifestations of the heisenberg uncertainty relations , which are among the most important principles of quantum mechanics .
squeezing , in particular that of the radiation field , has attracted a great deal of interest and been extensively studied both theoretically and experimentally over the past thirty years due to its practical applications in many optical devices as well as in the quantum information , e.g. quantum teleportation @xcite , power - recycled interferometer @xcite and phase - modulated signal recycled interferometer @xcite .
meanwhile , increased attention has also been devoted to the squeezing of the fluctuations of the atomic dipole variables @xcite owing to its potential applications in the high - resolution spectroscopy @xcite , the high - precision atomic fountain clock @xcite , the high - precision spin polarization measurements @xcite , etc .
moreover , the important connection of atomic dipole squeezing with the squeezing of the radiation field has been established @xcite ; atomic dipole squeezing is shown to produce reduced quantum fluctuations in the radiated field . in order to analyze the atomic dipole squeezing in the model under consideration
, we consider the two slowly varying hermitian quadrature operators @xmath122 in fact @xmath123 , and @xmath124 correspond to the dispersive and absorptive components of the amplitude of the atomic polarization , respectively .
they obey the commutation relation @xmath125 = \frac{i}{2}\hat \sigma _ z$ ] .
correspondingly , the heisenberg uncertainty relation is @xmath126 , where @xmath127 is the variance in the component @xmath128 of the atomic dipole .
the fluctuations in the component @xmath128 are said to be squeezed @xcite if the variance in @xmath129 satisfies the condition @xmath130 . since @xmath131
this condition may be written as @xmath132 the expectation values of the atomic operators @xmath133 and @xmath134 in the state @xmath66 of the atom - field system , given by ( 25 ) , are @xmath135 @xmath136 let us examine numerically the squeezing properties of the atomic dipole variable @xmath134 in the model under consideration . figures .
2(a)-2(c ) show the time evolution of @xmath137 , corresponding to the squeezing of the component @xmath134 , as a function of the scaled time @xmath107 , when we consider the same values of parameters taken in figs.1 . as it is seen , when the influence of the gravitational field increases , the dipole squeezing is weakened .
also , numerical analysis reveals that increasing of the nonlinearity ( increasing the value of the deformation parameter _ q _ ) seriously reduces the dipole squeezing .
the next quantity we consider is the atomic momentum diffusion . as a consequence of the atomic momentum diffusion ,
the atom experiences light - induced forces ( radiation force ) during its interaction with the radiation field .
the atomic momentum diffusion is given by @xmath138 by using equation(25 ) , we obtain @xmath139 ^ 2 \ } ^{1/2}. \end{array}\ ] ] now we turn our attention to examine numerically the dynamics of the atomic momentum diffusion .
the numerical results of the evolution of @xmath140 are shown in figs .
3(a)-3(c ) with the same corresponding data used in fig.1 . + figure 3(a ) illustrates the case when the gravitational influence is negligible , while figs . 3(b ) and 3(c ) display the cases when we consider the gravitational influence .
it is seen from fig .
3(a ) that , when the gravitational influence is negligible , the rabi - like oscillations can be identified . however , by increasing the gravitational influence ( figs .
3(b ) and 3(c ) ) the momentum diffusion oscillates so drastically that the phenomenon of collapse - revival is not so clear .
moreover , the atom can experience larger light - induced forces during its interaction with the radiation field , when the gravitational field increases .
further numerical analysis shows that , for both cases of @xmath117 and @xmath121 , with the increasing value of the deformation parameter _
q _ the momentum diffusion increases and it exhibits a chaotic - like behaviour .
one of the best - known nonclassical effects of the light field is sub - poissonian photon statistics .
the experiments on sub - poissonian statistics are concerned with the intensity or photon - number fluctuations of the electromagnetic field .
to understand the influence of gravity on sub - poissonian statistics of the cavity - field in the _
f_-deformed jc model , we examine the behaviour of the second - order correlation function for the cavity - field defined by @xcite @xmath141 ^ 2 } } } , \end{array}\ ] ] where the photon probability distribution function at time @xmath68 , @xmath142 , is given by @xmath143 the photon statistics is said to be sub - poissonian if @xmath144 and super - poissonian if @xmath145 . the value @xmath146 stands for poissonian statistics . in figs .
4(a)-4(c ) we have sketched the temporal evolution of @xmath147 as a function of the scaled time @xmath107 with the same data as taken before . as it is seen from these figures , the rabi - like oscillations in the temporal evolution of @xmath147 can be identified .
furthermore , one can see that the initial photon statistics of the cavity - field is only slightly distorted by the interaction with the atom .
figure 4(a ) shows that when the influence of the gravitational field is negligible the cavity - field exhibits sub - poissonian statistics in the course of interaction ( fig .
4(a ) ) . with increasing @xmath108
the sub - poissonian characteristic is suppressed and the cavity - field exhibits super - poissonian statistics ( figs.4(b),4(c ) ) .
also , numerical analysis reveals that by increasing the deformation parameter _
_ , the second - order correlation function shows chaotic - like oscillations and the super - poissonian behaviour of the cavity - field in the presence of gravity is strengthened .
squeezed states are shown by reduced quantum fluctuations in one quadrature of the field at the expense of the increased fluctuations in the other quadrature . in order to investigate the quadrature squeezing of the _ f_-deformed cavity - field in the presence of homogenous gravitational field
, we introduce the two slowly varying hermitian @xmath11-deformed quadrature components @xmath148 and @xmath149 defined , respectively , by @xmath150 where the operators @xmath151 and @xmath152are the _
f_-deformed annihilation and creation operators given by ( 2 ) . in the limiting case @xmath153 , the operators @xmath148 and @xmath149 reduce to the conventional ( nondeformed ) quadrature operators @xcite .
f_-deformed quadrature operators obey the commutation relation@xmath154=i/2((\hat{n}+1)f^2(\hat{n}+1)-\hat{n}f^2(\hat{n } ) ) , \end{array}\ ] ] and , as result , the variances @xmath155 satisfy the uncertainty relation @xmath156 a state of the field is said to be squeezed when one of the _ f_-deformed quadrature components @xmath148 and @xmath149 satisfies the relation @xmath157 the degree of squeezing can be measured by the squeezing parameter @xmath158 defined by @xmath159 which can be expressed in terms of the _ f_-deformed annihilation and creation operators of the field as follows @xmath160 where @xmath161 then , the condition for squeezing in the _
f_-deformed quadrature component can be simply written as @xmath162 .
now , we investigate the temporal behaviour of @xmath163 , which gives information on deformed quadrature squeezing of @xmath164 , when we consider the same data as used before .
numerical results are presented in figs.5(a)-5(c ) , where we have plotted @xmath163 versus @xmath107 .
figure 5(a ) shows that despite the fact that the cavity - field remains in the state quite close to the initial state , the _ f_-deformed quadrature components @xmath164 exhibits squeezing in the course of time evolution when the influence of the gravitational field is negligible .
however , in view of figs .
5(b ) and 5(c ) we observe that the gravitational field leads to the disappearance of quadrature squeezing completely .
besides , numerical analysis shows that by increasing the nonlinearity of the model ( increasing the value of the deformation parameter _ q _ ) the quadrature squeezing occurs in smaller amount for a short time in the very beginning of the interaction when the influence of gravity is negligible .
furthermore , with the increasing value of @xmath108 the quantum fluctuations in the quadrature component @xmath164 increase considerably .
we proposed an _ f_-deformed formalism to study the quantum statistical properties of nonlinear jc model , including possible forms of nonlinearities of both the field and the intensity - dependent atom - field coupling , in the presence of a classical gravitational field .
the work here extends previous studies in this context @xcite . to analyze the dynamical evolution of the atom - radiation system
, we presented a quantum treatment of the internal and external dynamics of the atom based on an alternative _
f_-deformed @xmath3 dynamical algebraic structure . by solving the schrdinger equation in the interaction picture
, we found the evolving state of the system by which the influence of the classical gravity field on the dynamical behaviour of the atom - radiation system was explored . whilst the model was quite general , we looked specifically at a special choice of the nonlinearity , i.e. , _
q_-nonlinearity . considering the cavity - field to be initially in a _ maths_-type _
q_-deformed coherent state and the two - level atom to be in a coherent superposition of its excited and ground states , we investigated the influence of gravity on the atomic population inversion , atomic dipole squeezing , atomic momentum diffusion , photon counting statistics , and deformed quadrature squeezing of the radiation field .
we found : ( 1 ) with the increasing value of the gravity - dependent parameter @xmath108 and nonlinearity ( deformation ) parameter _ q _ the rabi - like oscillations in the atomic population inversion are deteriorated .
also , the presence of the gravitational field causes more energy to be stored in the cavity - field while in the absence of gravitational influence , the deformation causes more energy to be stored in the atomic subsystem , ( 2 ) both the gravitational field and the nonlinearity associated with the _ f_-deformation cause the atomic dipole squeezing to be weakened , ( 3 ) both the gravitational influence and the _ f_-deformation cause the atom to experience larger light - induced forces during its interaction with the radiation field , ( 4 ) with increasing the parameters @xmath108 and _ q _ the sub - poissonian behaviour of the cavity field is suppressed and it exhibits super - poissonian statistics and ( 5 ) the deformed quadrature squeezing of the cavity - field decreases with increase of the gravity - dependent parameter @xmath108 and the deformation parameter _
q_. to make the model under consideration more realistic the atom decay and cavity - field decay should be taken into account .
we hope to report on such issues in a forthcoming paper .
the authors would like to express their gratitude to the referees , whose valuable comments have improved the paper .
they are also grateful to the office of graduate studies of the university of isfahan for their support .
10 natexlab#1#1bibnamefont # 1#1bibfnamefont # 1#1citenamefont # 1#1url # 1`#1`urlprefix e. t. jaynes and f. w. cummings , _ proc . ieee _ * 51 * , 89 ( 1963 ) .
j. h. eberly , n. b. narozhny , and j. j. sanchez - mondragon , _ phys .
lett . _ * 44 * , 1383 ( 1980 ) ; + h. i. yoo , j. j. sanchez - mondragon , and j. h. eberly , _
* 14 * , 1383 ( 1981 ) .
p. meystre and m. s. zubairy , _ phys .
a _ * 89 * , 390 ( 1982 ) ; + j. r. kuklinski and j. c. madajczyk , _ phys . rev .
a _ * 37 * , r3175 ( 1988 ) . s. m. barnett , _ opt .
commun . _ * 61 * , 432 ( 1982 ) ; + p. zhou and j. s. peng , _ phys .
a _ * 44 * , 3331 ( 1991 ) . s. j. d. phoenix and p. l. knight , _ ann .
( n.y ) _ * 186 * , 381 ( 1988 ) .
d. meschede , h. walther , and g. muller , _ phys .
lett . _ * 54 * , 551 ( 1985 ) .
p. bertet , et al .
lett . _ * 88 * , 143601(2002 ) .
t. meunier , et al . ,
* 94 * , 010401 ( 2005 ) .
s. osnaghi , et.al . ,
lett . _ * 87 * , 037902 ( 2001 ) ; b. buck and c. v. sukumar , _ phys . lett . a _ * 81 * , 132 ( 1981 ) .
v. buzek , _ phys .
a _ * 39 * , 3196 ( 1989 ) . m. chaichian , d. ellinas , and p. kulish , _ phys .
lett . _ * 65 * , 980 ( 1990 ) .
p. gora and g. jedrzejek , _ phys .
a _ * 45 * , 6816 ( 1992 ) .
v. buzek , _ j. mod
. phys . _ * 39 * , 949 ( 1992 ) .
a. joshi and s.v .
lawande , _ int .
b _ * 6 * , 3539 ( 1992 ) ; + v. bartzis , _ physica a _ * 180 * , 428 ( 1992 ) .
d. bonatsos , c. daskaloyannis , and g.a .
lalazissis , _ phys .
rev . a _ * 47 * , 3448 ( 1993 ) j. crnugelj , m. martinis , and v. mikuta - martinis , _ phys .
b _ * 318 * , 227(1993 ) ; + j. crnugelj , m. martinis , and v. mikuta - martinis , _ phys . lett .
a _ * 188 * , 347(1994 ) . b. deb and s. sen , _ phys .
a _ * 56 * , 2470(1997 ) . m. f. fang , _ physica a _ * 259 * , 193 ( 1998 ) .
m. h. naderi , m. soltanolkotabi , and r. roknizadeh , _ j. phys .
jpn . _ * 73 * , 2413 ( 2004 ) .
a. joshi and min xiao , _ opt . commun . _
* 232 * , 273 ( 2004 ) ; m. h. naderi , m. soltanolkotabi , and r. roknizadeh , _ eur . phys
. j. d _ * 32 * , 397 ( 2005 ) .
m. h. naderi and m. soltanolkotabi , _ eur .
j. d _ * 39 * , 471 ( 2006 ) .
j. g. muga and b. navarro _ phys .
a _ * 73 * , 022715 ( 2006 ) .
m. h. naderi , _ can
. j. phys . _ * 85 * , 1071(2007 ) .
xie , f. jia , and y - p .
yang , _ opt .
commun . _ * 273 * , 451 ( 2007 ) .
j. b. geloun , j. govaerts , and m. n. hounkonnou , _ j. math .
* 48 * , 032107(2007 ) .
a. romanelli , _ phys .
a _ * 80 * , 014302 ( 2009 ) . m. jimbo , _ lett .
phys . _ * 10 * , 63 ( 1985 ) .
j. wess and b. zumino , _ nucl .
b , proc . suppl . _
* 18 * , 302 ( 1991 ) .
d. bonatsos and c. daskaloyannis , _ phys .
a _ * 46 * , 75 ( 1992 ) .
r. j. mcdermott and a. i. solomon , _ j. phys .
* 27 * , l15 ( 1994 ) .
m. chaichian and p. kulish , _ phys .
b _ * 234 * , 72 ( 1990 ) .
m. arik and d. d. coon , _ j. math
. phys . _ * 17 * , 524 ( 1976 ) .
r. l. de matos filho , w. vogel , _ phys .
a _ * 54 * , 4560 ( 1996 ) ; + v. i. manko , g. marmo , e. c. g. sudarshan , and f. zaccaria , _ phys . scripta _ * 55 * , 528 ( 1997 ) .
y. wu and x. yang , _ phys .
a _ * 56 * , 2443 ( 1997 ) . c. baxter , m. babiker , and r. loudon , _ phys .
a _ * 47 * , 1278(1993 ) .
a. kastberg , w. d. philips , s. l. rolston , r. j. c. spreeuw , and p. s. jessen , _ phys .
* 74 * , 1542 ( 1995 ) .
m. mohammadi , m. h. naderi , and m. soltanolkotabi , _ j. phys .
a : math . theor .
_ * 40 * , 1377 ( 2007 ) .
i. bloch , t. w. hansch , and t. esslinger , _ nature ( london ) _ * 403 * , 166 ( 2000 ) . v. v. nesvizhevsky , et al .
, _ nature ( london ) _ * 415 * , 297 ( 2002 ) .
c. lammerzahl and c. j. borde , _ phys .
a _ * 203 * , 59 ( 1995 ) .
k. p. marzlin and j. audertsch , _ phys .
a _ * 53 * , 1004 ( 1995 ) .
m. mohammadi , m. h. naderi , and m. soltanolkotabi , _ j. phys .
a : math . gen . _ * 39 * , 11065 ( 2006 ) .
m. mohammadi , m. h. naderi , and m. soltanolkotabi , _ int .
phys_. * 47 * , 983 ( 2008 ) .
m. mohammadi , m. h. naderi , and m. soltanolkotabi , _ eur .
j. d _ * 47 * , 295 ( 2008 ) .
m. kozierowski and s. m. chumakov _ in coherence and statistics of photons and atoms _ , edited by j.perina ( john wiley sons , new york , 2001 ) chap.9 , pp.375 - 421 .
m. o. scully and m. s. zubairy , _ quantum optics _ ( cambridge university press , cambridge , 1997 ) .
j. perina , jr . and
j. perina , _ quantum statistics of nonlinear optical couplers _ ,
progress in optics , vol .
e. wolf , elsevier , amsterdam 2000 , p. 361 .
g. khitrova , h. m. gibbs , m. kira , s. w. koch , and a. sherer , _ nature phys . _
* 2 * , 81(2006 ) .
j. keeling and v. gurarie , _ phys . rev .
* 101 * , 033001 ( 2008 ) .
d. m. meekhof , c. monroe , b. e. king , w. m. itano , and d. j. wineland , _ phys .
lett . _ * 76 * , 1796 ( 1996 ) .
s. l. braunstein and h. j. kimble , _ phys .
rev . lett_. * 80 * , 869(1998 ) .
a. brillet , j. gea - banacloche , g. leuchs , c. n. man , and j. y. vinet , _ the detection of gravitational radiation _ , ed .
d. g. blair ( cambridge : cambridge university press , 1991 ) .
v. chickarmane , s. v. dhurandhar , t. c. ralph , m. gray , h - a .
bachor , and d. e. mcclelland , _ phys .
a _ * 57 * , 786(1998 ) .
d. f. walls and p. zoller , _ phys .
* 47 * , 709 ( 1981 ) ; + r. loudon , _ opt .
_ * 70 * , 109 ( 1989 ) .
m. kitagawa and m. ueda , _ phys .
a _ * 47 * , 5138 ( 1993 ) .
d. j. wineland , j. j. bollinger , w. m. itano , and d. j. heinzen , _ phys .
a _ * 50 * , 67(1994 ) . j. l. sorensen , j. hald , and e. s. polizik , _ phys .
lett . _ * 80 * , 3487 ( 1998 ) .
k. wodkiewicz , p. l. knight , s. j. buckle , and s. m. barnett , _ phys .
rev . a _ * 35 * , 2567 ( 1987 ) .
m. hillery , _ phys . rev .
a _ * 61 * , 022309 ( 2000 ) .
* figure1 . * time evolution of the atomic population inversion @xmath106 versus the scaled time @xmath107 .
here we have set @xmath165 and @xmath109 ( 1a ) for @xmath117 ( 1b ) for @xmath166(1c ) for @xmath167 .
+ + * figure2 .
* time evolution of @xmath137 , corresponding to the squeezing of the component @xmath168 , versus the scaled time @xmath107 , with the same corresponding data used in figure 1:(1a ) for @xmath117 ( 1b ) for @xmath166(1c ) for @xmath167 . + + * figure 3 .
* time evolution of the atomic momentum diffusion @xmath140 versus the scaled time @xmath107 , with the same corresponding data used in figure 1:(1a ) for @xmath117 ( 1b ) for @xmath166(1c ) for @xmath167 .
+ + * figure 4 .
* time evolution of the second - order correlation function @xmath169 versus the scaled time @xmath107 , with the same corresponding data used in figure 1:(1a ) for @xmath117 ( 1b ) for @xmath166(1c ) for @xmath167 .
+ + * figure 5 .
* time evolution of the squeezing parameter @xmath163 , corresponding to the squeezing of the deformed quadrature component @xmath170versus the scaled time @xmath107 , with the same corresponding data used in figure 1:(1a ) for @xmath117 ( 1b ) for @xmath166(1c ) for @xmath167 . | * abstract * + in this paper , we investigate the effects of a classical gravitational field on the dynamical behaviour of nonlinear atom - field interaction within the framework of the _ f_-deformed jaynes - cummings model . for this purpose ,
we first introduce a set of new atomic operators obeying an _ f_-deformed _
su(2 ) _ algebraic structure to derive an effective hamiltonian for the system under consideration . then by solving the schrdinger equation in the interaction picture and considering certain initial quantum states for the atomic and radiation subsystems , we analyze the influence of gravity on the temporal evolution of the atomic population inversion , atomic dipole squeezing , atomic momentum diffusion , photon counting statistics , and deformed quadrature squeezing of the radiation field . |
rotation curves are the principal tool to derive the axisymmetric distribution of mass in spiral galaxies in the first - order approximation
. rotation curves of galaxies in the disk and outer regions have been obtained based on optical and hi - line spectroscopy ( rubin et al 1980 , 1982 ; bosma 1981 ; clemens 1985 ; mathewson et al 1996 ; persic and salucci 1995 ; persic et al 1996 ; honma and sofue 1997 ) .
these rotation curves have been used to estimate the mass distribution in the disk and halo , particularly for the dark halo , using a model - potential fits method ( e.g. , kent 1987 , 1992 ) .
persic et al ( 1996 ) have extensively studied the universal characteristics of rotation curves in the outermost regions .
however , the inner rotation curves have been not thoroughly investigated yet in sufficient accuracy , not only because the concern in these studies has been on the distribution of mass in the outermost regions , but also because of the difficulty in deriving the central rotation velocities .
the difficulty in measuring central rotation curves is mainly due to the lack of hi gas in the central regions , as well as due to contamination of bright bulge light , when photographic plates had been used . in our recent papers ,
we have stressed that , in order to derive central rotation curves , co molecular lines are most convenient because of the high concentration of molecular gas in the centers of many galaxies , the high angular and velocity resolutions in co observations , and the negligible extinction even toward the nuclear dusty disk .
we have thus used high - resolution co - line data to obtain well - sampled inner rotation curves for nearby galaxies ( sofue 1996 , 1997 , sofue et al 1997 , 1998 : papers i to iv ) .
recent ccd line spectroscopy has also made available to us accurate rotation curves for the inner regions , because of the large - dynamic range , and the more precise subtraction of bulge continuum light for digital images , in so far as the extinction is not very large : the obtained optical position - velocity ( pv ) diagrams often show high - velocity central components ( rubin et al 1997 ; sofue et al 1998 ; bertola et al 1998 ) .
however , it is not yet known whether the central steep rise of rotation velocity is universal , how it is correlated to other properties of galaxies , such as the hubble types and central activity . in this paper
, we present high - accuracy rotation curves for sb , sbb , sc and sbc galaxies , and discuss their general characteristics , particularly for central rotation curves . we also discuss their dependency on morphological types , activity , and peculiarity .
individual rotation curves are shown in the appendix , and machine - readable data are available by contacting the first author .
the rotation curve data , from the central to outer regions in high accuracy , will be used in deriving the detailed mass and mass - to - luminosity ratios in a separate paper ( takamiya and sofue 1999a ) .
the rotation curve of the milky way galaxy is shown in figure 1a , as reproduced from the literature ( clemens 1985 ; honma and sofue 1997 ) . recently
, it has been shown that the velocity dispersion of stars within the central 10 pc increases toward the center , indicating the existence of a massive black hole ( genzel et al 1997 , 1998 ; ghez et al 1998 ) . assuming the existence of a black hole of @xmath0 , we have calculated a corresponding rotation curve , and combined it with the existing co data to obtain a rotation curve of the milky way as shown in figure 1b in logarithmic plot . besides the galaxy
, evidence for nuclear massive black holes has been accumulating for many other galaxies ( miyoshi et al 1995 ; richstone , et al .
1998 ) , suggesting similar central rotation curves for these galaxies ( see section 3.3 ) .
thus , the rotation curve of our galaxy can be described as having \(a ) a high - density core , including the massive black hole , which causes a non - zero velocity very close to the center .
\(b ) a steep rise within the central 100 pc .
\(c ) a maximum at radius about 300 pc , followed by a decline to a minimum at 2 kpc .
\(d ) a gradual rise from 2 kpc to the disk maximum at 6 kpc .
\(e ) a nearly flat outer region with a dip at 8 kpc , followed by a second maximum at 15 kpc , followed further by an outermost keplerian decline . -
fig . 1 except for the milky way
, it has been widely believed that the central rotation curves of most galaxies behave in a rigid - body fashion .
these rotation curves appear to behave differently from that of the milky way s rotation in the inner disk and bulge regions .
the question may arise then as to whether the milky way really is an exception , or whether a similar rotation property is not showing up in the currently published rotation curves of galaxies .
we have performed high - resolution co - line observations of galaxies to obtain pv diagrams , and combined them with existing hi and optical rotation curves .
we have also obtained ccd spectroscopy in the and [ nii ] line emissions of the central regions of galaxies .
we have applied the envelope - tracing method to derive rotation curves from pv diagrams . in figure 2a
we show the thus obtained most - completely sampled rotation curves for all sample galaxies ( papers i - iv ) , and in figure 2b the same but for the central 5 kpc .
figure 2c shows some rotation curves plotted against radii normalized by the scale length radius of the exponential disk .
in deriving rotation curves , we have applied the envelope - tracing method , which traces the terminal velocities in position - velocity diagrams along the major axes , which is described in detail in papers i and ii . this method has an advantage to estimate the rotation velocity in the central regions more accurately than those used in the current analyses ( e.g. , warner et al 1972 ; rubin et al 1980 ; persic et al 1996 ) , which gave intensity - weighted velocities .
( see also section 4.2 . ) in figures 3a , b and c , we show rotation curves for sb , sc , and barred ( sbb and sbc ) galaxies , respectively .
figure 4 shows rotation curves for peculiar galaxies and interacting galaxies .
individual rotation curves are published in the literature as above with detailed discussion , and displayed in figure a1 of appendix .
parameters for individual galaxies are given in table 1 .
2 fig . 3
all sb galaxies in figure 3a have rotation curves with a very steep rise in the central 100 - 200 pc region , often associated with a peak at radii @xmath1 pc . the rotation velocity
, then , declines to a minimum at @xmath2 kpc , and is followed by a gradual rise to a broad maximum at @xmath3 kpc , corresponding to the disk .
the outermost parts are usually flat , indicating the massive dark halo .
some galaxies like the milky way show a declining outer rotation ( honma and sofue 1997a , b ) .
thus , the rotation curves for sb galaxies are essentially the same as that of the milky way galaxy .
sc galaxies tend to have slower velocities than sb , and the rotation velocities are more spread from @xmath4 to @xmath5 among the sample galaxies .
massive sc galaxies show a steep nuclear rise similar to sb s , while less - massive galaxies have a more gentle rise .
they also have a flat rotation until their outer edges . in figure 3
we compare barred galaxies with non - barred galaxies .
there appears to be no particular difference in their general properties : the rotation properties of the barred galaxies are almost the same as those for non - barred galaxies of sc and sb types . however , barred galaxies tend to show a larger - amplitude velocity variation with radius of about @xmath6 within the main disk at @xmath7 kpc
. the large velocity variation may be due to barred potential of length of several kpc . in order to clarify this and investigate
the non - circular motions , two - dimensional velocity fields are necessary , which is , however , out of the scope of this paper , and remains as a future subject . on the other hand ,
normal galaxies usually show velocity variation of about @xmath8 mainly caused by spiral arms , except for a few cases .
rotation curves for irregular galaxies ngc 660 , ngc 3034 , ngc 4631 and ngc 4945 are shown in figure 4a .
ngc 660 is a polar - ring galaxy with a warped galactic disk , while the rotation curve from the disk to outer polar ring behaves as if they are continuous structures .
ngc 3034 ( m82 ) shows a very exceptional behavior of rotation : it has a steep nuclear rise as usual , but decreases after the nuclear peak , obeying the keplerian law .
this may have occurred by a strong tidal truncation of the disk , by a close encounter with m81 ( sofue 1998 ) .
ngc 4631 is an interacting dwarf galaxy with peculiar morphology , showing a rigid - body increase of rotation velocity .
since this galaxy is edge on , it is not clear , either if the co gas is lacking indeed in the center , or it results from a true rigid - body rotation .
ngc 4945 shows peculiar dark lanes and patches , but shows a quite normal rotation : steep nuclear rise and flat disk rotation .
its peculiar distribution of interstellar gas is probably not due to its global dynamical characteristics , but due to a local phenomenon in the gas disk .
the interacting galaxy ngc 5194 ( m51 ) shows a very peculiar rotation curve ( fig .
a1 ) , which declines more rapidly than the keplerian law at @xmath9 kpc .
this may be due to variation of the inclination angle with the radius , or warping .
in fact , this galaxy is nearly face - on ( @xmath10 ) , but a slight warp would cause a large error in deriving the rotation velocity : if the galaxy s outer disk at 12 kpc has an inclination as small as @xmath11 , such an apparently steep decrease would be observed even for a flat rotation .
fig.4 our sample includes galaxies having various activities , such as starbursts ( ngc 253 , ngc 1808 , ngc 3034 ) , seyferts ( ngc 1068 , ngc 1097 ) , liners ( ngc 3521 , ngc 4569 , ngc 7331 ) , and nuclear jets ( ngc 3079 ) . in figure
4c we plot rotation curves for these galaxies .
the global rotation and mass distribution appear rather normal in these active galaxies , and no peculiar behavior is found within our resolution .
this implies that such activities are triggered by a more local and secondary cause than by a global dynamical mass distribution .
an exception is the starburst galaxy ngc 3034 ( m82 ) , which shows a usual nuclear rise but is followed by a keplerian declining rotation , indicating a tidal truncation of the disk ( sofue 1998 ) , as mentioned in the previous subsection .
in figure 5 we show mean rotation curves , which have been obtained by averaging rotation velocities for three mass classes : massive galaxies with a maximum disk velocity greater than 250 ; galaxies with maximum disk velocities between 200 and 250 ; and less - massive galaxies with velocities less than 200 . in figure 5 , we also plot the universal rotation curves ( urc ) formulated by persic et al ( 1996 ) .
the observed mean rotation curves at @xmath12 kpc are well fitted by the urc . since the urc has been derived specifically for studying massive halos , the disk and central rotation curves may not be well reproduced .
in fact , the central rotation curve obtained in this study is much steeper than the urc
. this may be partly due to the difference in the method to derive the rotation curve : rotation curves in the previous studies have been usually derived from the intensity peak in the pv diagrams . on the other hand , in the present study we used the envelope - tracing method , which is thought to be more reliable to trace the central rotation curve ( sofue 1996 ; takamiya and sofue 1999b ) . a relation discussion of rotation curves and position - velocity diagrams will be given in subsection 4.2 .
the observed rotation curves can be classified into the following three types , according to their behavior in the central regions .
figure 6a to 6c show the classified rotation curves .
_ central peak type _ : rotation velocity attains a sharp maximum near the center at @xmath13 pc , followed by a dip at @xmath14 kpc , then by a broad maximum of the disk component .
examples are the milky way , ngc 891 , ngc 3079 , ngc 5907 , and ngc 6946 .
_ no central - peak type _
: the rotation curve rises steeply at the center , followed immediately by a flat part .
examples are ngc 224 , ngc 253 , ic 342 , ngc 5194 , and ngc 5055 .
_ rigid - body type _ : the rotation velocity increases mildly from the center in a rigid - body fashion within the central 1 kpc .
this type is rather the exception , and is found in less - massive sc - type galaxies , such as ngc 598 , ngc 2403 , ngc 3198 , and ngc 4631 .
this tendency has been already noticed by casertano and van gorkom ( 1991 ) .
fig . 6 since the dynamical structure of a galaxy varies with the radius rapidly toward the center , an alternative plot , such as in logarithm , of rotation curves would help to overview the innermost kinematics .
in fact , the logarithmic plot in figure 1b has demonstrated its convenience for discussing the central mass distribution , including the black hole of our galaxy . in figure 7 , we plot the same rotation curves as in figure 2a in logarithmic scaling .
the central 1 kpc regions are better presented in this figure .
however , resolutions of our data are not sufficient to express rotation characteristics in the very center , particular within 100 pc for many galaxies .
although we must be careful in looking at this figure in the sense that the resolution is limited in the central few hundred pc , we may safely argue that high - mass galaxies show almost constant rotation velocities from the center to outer edge , in so far as they are presented in such a logarithmic plot .
on the other hand , lower - mass galaxies show decreasing rotation toward the center in the central 1 kpc regions .
however , we must also notice that the declining rotation toward the center might be caused by insufficient angular resolution , with which position - velocity diagrams often miss the central steep rise .
fig . 7 a logarithmic rotation curve is particularly useful for such cases with central massive black holes .
figure 7b shows rotation curves of four galaxies from our sample , for which the existence of massive black holes is evident ( richstone et al 1998 ) : ngc 4258 ( miyoshi et al 1995 ) , milky way ( genzel et al 1997 ; ghez et al 1998 ) , ngc 224 ( magorrian et al 1998 ) and ngc 4945 ( greenhill et al 1997 ) .
here , equivalent rotation velocities corresponding to the black hole masses are plotted by tilted straight lines , obeying @xmath15 . except for the milky way , they are connected to known rotation curves by horizontal straight lines in the regions where detailed observations are lacking . in these galaxies ,
the rotation velocity never declines to zero at the center .
maximum rotation velocity is an indicator of total mass of a galaxy ( e.g. persic et al 1996 ) .
the disk mass is also related to the optical scale radius ( @xmath16 ) , because the mass - to - luminosity ratio would not vary significantly inside the disk .
it is interesting to see how the scale radius of a disk is correlated with the maximum velocity of rotation , which will occur at @xmath17 for an exponential disk as is observed in fig .
2 . in figure 8
we plot rotation velocities at @xmath18 and @xmath17 against the scale radius , @xmath16 , from our sample galaxies , where open circles are sb , filled circles sc , and those with triangles are barred galaxies . although the correlation between velocity and scale radius is not tight , there appear to exist two types of rough correlation s : one group contains galaxies with rotation velocities linearly correlated with the scale radius , approximately fitted by the dashed line .
another group comprises galaxies with largely - scattered velocities but with scale radii being about constant at @xmath19 kpc .
no particular trend , such as larger scatter , for barred galaxies is found in this plot .
the extremely high frequency of massive galaxies showing the steep nuclear rise indicates that the high velocity is not due to a particular view of non - circular motion by chance .
note that the probability of looking at a bar end - on is much smaller than that of viewing one side - on , which should result in a larger probability for apparently slower rotation than circular velocity at a given radius , and therefore , masses estimated from the circular assumption would be even underestimated , if they contain a bar . in
so far as our sample galaxies are concerned , the steep nuclear rise of rotation velocity is a universal property for massive sb and sc galaxies , regardless of the existence of a bar and morphological peculiarities .
however , less - massive galaxies tend to show a rigid - body rise .
we stress that the rotation velocities in many well - resolved galaxies do not decline to zero at the nucleus .
this indicates that the mass density increases toward the nucleus more rapidly than expected from exponential or de vaucouleur laws .
we mention that the widely adopted zero - velocity in the center may be merely due to a custom of drawing a rotation curve by linking positive and negative velocities from the opposite sides of the nucleus .
we point out that nearer galaxies with higher effective resolution in our sample tend to show steeper nuclear rise .
this suggests that an even steeper central rise would be observed in more distant galaxies , if they were observed at higher resolution . in order to examine how central high velocities would appear in an observed pv diagram with finite resolution
, we have simulated pv diagrams from assumed rotation curves and a model distribution of interstellar gas .
figure 9a shows a case in which the rotation velocity is assumed to be constant , so that the velocity does not decline to zero at the center , and the interstellar gas is centrally peaked .
however , the calculated pv diagram shows a rigid - body behavior near the center , and if we trace the intensity peak on this pv diagram , the velocity at the center would be measured to be zero . hence , in so far as a rotation curve constructed from a pv diagram is concerned , we should not take its central rigid body - like behavior strictly : the true velocity may be much higher , or may even start from a finite value near a central black hole .
figure 9b shows a case with a more realistic rotation curve , comprising four components : a central compact core , bulge , disk and massive halo , each expressed by a plummer potential . in the calculated pv diagram , however , the central steep rise and the peaks due to the core and bulge are hardly recognized .
this simulation demonstrates that tracing peak - intensity loci in observed pv diagrams could miss possible central high velocities .
we , thus , conclude that the central rotation curves derived from observed pv diagrams generally give _
lower limits _ to the rotation velocities .
in fact , high - resolution central pv diagrams observed for some spiral galaxies by bertola et al ( 1998 ) have indicated velocities much higher than those currently supposed , indicative of massive central black holes . detailed discussion of pv diagrams and a method to derive more reliable rotation curves will be given in a separate paper ( takamiya and sofue 1999b ) .
we present rotation curves for individual galaxies in figure a1 .
machine - readable data in the form of tables are available by contacting the first author .
table a1 lists the parameters of the galaxies .
most - completely - sampled rotation curves of sb and sc galaxies obtained by using co , and hi - line data .
( a ) all galaxies ; ( b ) all galaxies for central 5 kpc ; ( c ) rotation curves with radii normalized by scale radius of exponential disk . fig .
5 . mean rotation curves for galaxies with disk rotation velocities higher than 250 , between 200 and 250 , and lower than 200 .
universal rotation curves ( urc ) formulated by persic et al ( 1996 ) are shown by dashed lines for several values of galaxy luminosities . fig .
simulation of position - velocity diagrams from assumed rotation curves and gas distribution : ( a ) rotation velocity is assumed to be constant , and therefore , the central velocity is finite .
( b ) four mass components ( core , bulge , disk and halo ) are assumed , having a very sharp central rise of rotation velocity . note that the high velocities in the center and steep nuclear rise are hardly detected in pv diagrams with finite resolution .
this simple simulation suggests that observed pv diagrams often miss a massive compact component in the center . | we present high - resolution central - to - outer rotation curves for sb , sbb , sc , and sbc galaxies .
we discuss their general characteristics , particularly their central behavior , as well as dependencies on morphological types , activity , and peculiarity .
the rotation curves generally show a steep nuclear rise , and high - velocity central rotation , followed by a broad maximum in the disk and then by a flat rotation due to the massive halo
. since the central high velocity and steep rise are common to all massive galaxies , it can not be due to non - circular motions .
disk rotation curves of barred galaxies show larger dispersion than those of normal galaxies which is probably due to non - circular motions .
interacting galaxies show often perturbed outer rotation curves , while their central rotation shows no particular peculiarity . also , central activities , such as starbursts and agn , appear to show no particular correlation with the property of rotation curves
. this would suggest that the central activities are triggered by a more local effect than the global dynamical property .
keywords : galaxies : kinematics galaxies : rotation curve galaxies : mass distribution = -0.5 cm = -2 mm = -2 mm |
the blume
capel ( bc ) model @xcite is a spin1 ising model with single
ion anisotropy that has been introduced as a model for magnetic systems and then applied to multicomponent fluids @xcite and extended to the blume emery griffiths model @xcite for he@xmath0he@xmath1 mixtures .
it is not exactly solvable in more than one dimension , but it is has been studied over infinite @xmath2dimensional lattices by means of many different approximate techniques , and its phase diagram is well known . in recent years ,
when the theory of surface critical phenomena started developing , some attention has been devoted to the study of the bc model over semi
infinite lattices , with modified surface couplings . in particular benyoussef , boccara and saber @xcite have determined the phase diagram in the mean field approximation , reporting four possible topologies at fixed bulk / surface coupling ratios , while benyoussef , boccara and el bouziani @xcite have done a similar analysis using a real space renormalization group transformation .
other works , referring to particular regions of the phase space , are those by peliti and leibler @xcite and crisanti and peliti @xcite ( real space renormalization group ) , jiang and giri @xcite ( mean field approximation ) , tamura @xcite ( effective field approximation ) and the present authors @xcite ( cvm , low temperature expansion ) .
all these results show that , as in the case of the spin1/2 ising model @xcite , it is possible to have a phase with ordered surface and disordered bulk , which is separated from the completely ordered phase by the so called _ extraordinary _ transition and from the completely disordered phase by the _ surface _ transition .
when such a phase is absent , the transition between the completely ordered and the completely disordered phase is named _ ordinary _ and the meeting point of the lines of these three kinds of phase transitions is said _ special _ , and is generally a multicritical point . while it is well established that the ordinary transition can be either first or second order , the possibility of a first order surface or extraordinary transition has been ruled out by benyoussef , boccara and el bouziani @xcite and tamura @xcite , but not by the other authors . in the present paper , which can be regarded as a refinement of a previous work @xcite
, we apply the pair approximation of the cluster variation method ( cvm ) @xcite to the blume capel model defined on a semi infinite simple cubic lattice with a ( 100 ) free surface . the dependence of the local quantities ( order parameters , energy density , and so on ) on the distance from the free surface is taken into account by dividing the lattice into an infinite set of layers parallel to the surface and then treating the @xmath3th layer ( we will use @xmath4 ) as if it was bulk .
our results confirm the existence of first order extraordinary and surface transitions ( for the surface one we obtain very good agreement with a previous low temperature expansion analysis ) .
moreover , we show clearly that another first order transition can occur at the surface in presence of an ordered bulk , which separates two phases with large and small values of the surface order parameters , the latter being a new phase .
when this new transition is included in the phase diagram , the classification given by benyoussef , boccara and saber @xcite is completed with the addition of two new topologies .
the paper is organized as follows : in section ii we define the model and determine its ground state ; in section iii we give the pair cvm free energy functional and describe the nested iteration procedure we use to find its minimum ; in section iv we describe our results , giving the complete classification of the phase diagram topologies and the behaviour of the order parameters and their profiles in the new phase . finally , in section
v , we briefly summarize and discuss our results .
the bc model on a semi
infinite lattice @xmath5 with a free surface denoted by @xmath6 is defined by the ( reduced ) hamiltonian @xmath7 where @xmath8 are site labels , @xmath9 is the @xmath10component of a spin1 operator at site @xmath11 , @xmath12 and @xmath13 denote the bulk and surface nearest neighbors ( reduced ) interactions ( assumed positive ) , respectively , @xmath14 and @xmath15 denote the bulk and surface single ion ( reduced ) anisotropy and @xmath16 , with @xmath17 boltzmann s constant and @xmath18 absolute temperature ( from now on , we will express the temperature in units of the bulk interaction , by setting @xmath19 ) .
furthermore , we introduce the coupling ratios @xmath20 and @xmath21 .
as it is often the case with classical discrete models , we can easily determine the ground state of our model by looking for the lowest energy configurations .
this analysis is very important since it drives all subsequent considerations . for the bulk ,
as is well known , we have a state with broken spin
flip symmetry for @xmath22 , given by e.g. @xmath23 for all bulk spins , and a state with unbroken symmetry , @xmath24 , for @xmath25 .
these two states are associated with ordered and disordered bulk phases , respectively , at finite temperature . at the surface
the situation is slightly more complex .
let us define the states lo and ho by @xmath26 and @xmath27 , respectively .
the corresponding energies per site will be @xmath28 and @xmath29 , where @xmath30 in the bulk disordered phase and @xmath31 in the bulk ordered phase ( @xmath32 is but the bulk order parameter @xmath33 at zero temperature ) .
we can now distinguish the following cases @xcite : \a ) when @xmath34 there is a surface transition at @xmath35 , in presence of an ordered bulk , and the states lo and ho are associated , at finite temperature , to two phases with ordered surface , but with different values of the surface order parameters : the phase ho ( we use the same symbols for the phases and the corresponding pure states at @xmath36 ) has large order parameters and is the usual ordered phase , while lo is a new phase with small order parameters ; this situation has not been considered before , except in the limit @xmath37 , which yields the spin1/2 ising model with a spin1 free surface studied by kaneyoshi @xcite , in which a first order surface transition in the presence of ordered bulk is obtained , but neither studied in detail nor reported in the phase diagram ; \b ) when @xmath38 there is a surface transition at @xmath39 , in correspondence of the bulk transition , and the states lo and ho are associated to phases with disordered and ordered surface , respectively ; \c ) when @xmath40 there is a surface transition at @xmath41 , in presence of a disordered bulk , and the states lo and ho have the same meaning as in the previous case .
in order to study the finite temperature properties of the model , we introduce now the pair approximation of the cluster variation method .
the cvm is based on an approximate expression of the entropy of the model as a sum of contributions by all the elements of a set @xmath42 consisting of certain maximal clusters and all their subclusters . in the pair approximation
, one takes as maximal clusters all nearest neighbor pairs , and thus the only subclusters to be considered are the sites .
to each cluster @xmath43 is associated a cluster ( reduced ) entropy given by @xmath44 where @xmath45 is a cluster density matrix , to be determined minimizing the free energy .
the total reduced entropy is then approximated by @xmath46 where the coefficients @xmath47 obey @xcite @xmath48 and @xmath49 represents the multiplicity ( number of clusters per site ) of the cluster @xmath43 .
since we have to deal with local quantities ( density matrix elements , in our formulation ) which depend on the distance from the surface , we divide our lattice , as mentioned in the introduction , into two dimensional layers parallel to the surface , labeled by an integer @xmath50 , @xmath51 being the surface layer . furthermore , to avoid working with an infinite number of variables
, we will treat the @xmath3th layer ( typically we will choose @xmath52 ) as if it was bulk , by means of suitable constraints , and the bulk itself will be considered in the cvm pair approximation , too , using the results obtained in @xcite .
following @xcite , we introduce the density matrices @xmath53 for the pairs having both sites in the @xmath50th layer , @xmath54 for those pairs having spin @xmath55 in layer @xmath50 and spin @xmath56 in layer @xmath57 , and @xmath58 for sites in layer @xmath50 .
the pair density matrices must satisfy the obvious constraints @xmath59 which ensure that the site density matrices can be properly defined , e.g. by @xmath60 .
\label{rho1n}\ ] ] finally , in order to make the @xmath3th layer represent the bulk in an effective way , we will treat the corresponding pair density matrix as a constant , defined by @xmath61 where @xmath62 is the normalized bulk pair density matrix given by the cvm pair approximation @xcite .
notice that the normalization of @xmath62 and the constraints eq .
[ compat ] imply the normalization of all the density matrices we use . using the coefficients @xmath47 and @xmath49 reported in @xcite one readily obtains the ( reduced ) free energy density @xmath63 \nonumber \\ & & + \sum_{n=1}^{n-1 } \left [ 2 { \rm tr } ( \rho_2^{(n ) } \ln \rho_2^{(n ) } ) + { \rm tr } ( \rho_2^{(n , n+1 ) } \ln \rho_2^{(n , n+1 ) } ) + a_1^{(n ) } { \rm tr } ( \rho_1^{(n ) } \ln \rho_1^{(n ) } ) \right ] \nonumber \\ & & + \sum_{s_1 } \sum_{n=1}^{n-1 } \lambda_+^{(n)}(s_1 ) \left [ \sum_{s_2 } \rho_2^{(n , n+1 ) } ( s_1,s_2 ) - \rho_1^{(n)}(s_1 ) \right ] \nonumber \\ & & + \sum_{s_1 } \sum_{n=2}^{n } \lambda_-^{(n)}(s_1 ) \left [ \sum_{s_2 } \rho_2^{(n-1,n ) } ( s_2,s_1 ) - \rho_1^{(n)}(s_1 ) \right ] , \label{free}\end{aligned}\ ] ] where @xmath64 are lagrange multipliers , @xmath65 and @xmath66 for @xmath67 . minimizing this
functional corresponds to solving the following equations , in the form of the natural iteration method @xcite , @xmath68 \times \nonumber \\ & & \times \exp\left [ j_{\rm s } s_1 s_2 - \frac{\delta_{\rm s}}{4}(s_1 ^ 2 + s_2 ^ 2 ) \right ] \rho_1^{(1)}(s_1)\rho_1^{(1)}(s_2 ) \nonumber \\
\rho_2^{(n>1)}(s_1,s_2 ) & = & \exp \left [ \frac{1}{4 } \left ( \lambda_+^{(n)}(s_1 ) + \lambda_+^{(n)}(s_2 ) + \lambda_-^{(n)}(s_1 ) + \lambda_-^{(n)}(s_2 ) \right ) \right ] \times \nonumber \\ & & \times \exp\left [ j s_1 s_2 - \frac{\delta}{4}(s_1 ^ 2 + s_2 ^ 2 ) \right ] \cdot \left [ \rho_1^{(n)}(s_1)\rho_1^{(n)}(s_2 ) \right]^{5/4 } , \nonumber \\
\rho_2^{(n , n+1)}(s_1,s_2 ) & = & \exp \left [ - \lambda_+^{(n)}(s_1 ) - \lambda_-^{(n+1)}(s_2 ) \right ] \exp \left ( j s_1 s_2 \right ) , \label{nim}\end{aligned}\ ] ] together with the minor iteration equations @xcite @xmath69_{r+1 } = \left [ \lambda_+^{(n ) } ( s_1 ) \right]_r + \frac{1}{2 } \ln \frac{\displaystyle\sum_{s_2 } \rho_2^{(n , n+1)}(s_1,s_2 ) } { \rho_1^{(n)}(s_1 ) } \nonumber \\ \left [ \lambda_-^{(n ) } ( s_1 ) \right]_{r+1 } = \left [ \lambda_-^{(n ) } ( s_1 ) \right]_r + \frac{1}{2 } \ln \frac{\displaystyle\sum_{s_2 } \rho_2^{(n-1,n)}(s_2,s_1 ) } { \rho_1^{(n)}(s_1 ) } , \label{mnr}\end{aligned}\ ] ] for the lagrange multipliers , where @xmath70 and @xmath71 denote successive steps of the iteration procedure .
the method of solution can be described as follows : 0 .
: : initialize all pair density matrices with a rough estimate of the solution ; 1 .
: : determine the site density matrices according to eq .
[ rho1n ] ; 2 .
: : recalculate the pair density matrices according to eq .
[ nim ] with all lagrange multipliers set to 0 ; 3 .
: : solve eq .
[ mnr ] for the lagrange multipliers by simple iteration ; 4 .
: : determine the pair density matrices using eq .
[ nim ] with the values of the lagrange multipliers obtained in step 3 ; 5 .
: : iterate steps 14 until the desired precision is reached . recalling
that , in region of the phase space close enough to first order transition lines , our free energy should have different minima corresponding to stable and metastable phases , one has to repeat the above procedure with different initializations , corresponding to the different phase , and then choose the solution of minimum free energy .
once the solution is found one readily obtains the order parameters @xmath72 and , in an analogous way , all the correlation functions .
by means of the nested iteration procedure described in the previous section , we have determined the phase diagram of our model in the plane @xmath73 , for fixed values of @xmath74 and @xmath75 .
we have obtained six main topologies , reported in fig .
1 , extending thereby the classification by benyoussef , boccara and saber @xcite , in which cases ( a ) and ( d ) were missing .
these topologies are identified by specifying the ground state , according to the classification given in section ii , and the result in the ising limit @xmath76 .
recalling the known results about the critical behaviour of the semi infinite ising model @xcite one realizes that in this limit there are two possibilities : for @xmath77 the model exhibits an ordinary transition at the ising bulk critical temperature , while for @xmath78 it undergoes an extraordinary transition at the bulk critical temperature and a surface transition at a higher temperature .
the best estimate for @xmath79 is about 0.66 @xcite . in fig . 1 ,
case ( a ) corresponds to @xmath77 and @xmath34 , case ( b ) to @xmath80 and @xmath38 , case ( c ) to @xmath77 and @xmath40 , case ( d ) to @xmath78 and @xmath34 , case ( e ) to @xmath81 and @xmath38 and , finally , case ( f ) to @xmath78 and @xmath40 .
notice that almost all of the above topologies , with the exception of cases ( b ) and ( f ) , could be further divided into subtopologies based on the order of surface and bulk transitions at the special point @xmath82 and on the eventual coalescence of that point with the critical point @xmath83 .
this is exemplified in fig .
2 , which refers to cases ( a ) and ( d ) of fig . 1 .
it is also interesting to look at the behaviour of the order parameters in the cases ( a ) and ( d ) of fig . 1 , where the new lo phase appears . in figs .
3a and 3b , which refer to case ( a ) , we see that the surface layer order parameters increase from zero at low temperatures ( it can be shown exactly that at low temperatures they behave as @xmath84 ) , while the order parameters of the inner layers behave nearly as the bulk ones . furthermore , in figs . 4a and 4b , the effect of the new surface phase transition is clearly shown .
let us now turn our attention to the first order surface transition which occurs in presence of a disordered bulk , which existence , questioned by previous works @xcite , is clearly confirmed in our approximation ( fig . 1 , cases ( c ) and ( f ) ) . in fig
5 we have compared , for one choice of @xmath74 and @xmath75 , our results with those obtained by mean field approximation @xcite and low temperature expansion @xcite .
it is clearly seen that our approximation is a marked improvement with respect to the simple mean field approximation , and the very good agreement with the low temperature expansion results is a strong argument in favour of the existence of such a transition . finally , we have not found any reentrant phenomenon in the phase diagram , contrarily to previous results @xcite .
this discrepancy should be due to the very rough approximation used in @xcite , where only one layer above the bulk was considered and , a factor even more important , the method of solution adopted did not satisfy the first of the constraints eq .
[ compat ] in the case @xmath85 ( = 2 , in that case ) .
we have determined the phase diagram of the semi infinite blume capel model using the pair approximation of the cluster variation method .
our analysis has shown , at fixed @xmath74 and @xmath75 , six main topologies , two of which are completely new because of the appearance of a new ordered phase and , consequently , of a new surface first order phase transition in presence of an ordered bulk .
this new phase has been studied in some detail and typical behaviours of the order parameters profiles have been reported .
we have also given an example of how these main topologies may be slightly modified by varying @xmath74 and @xmath75 .
a question concerning the existence of the surface first order transition in the presence of disordered bulk has been addressed and clarified by comparing our results with previous results obtained by a low temperature expansion and , finally , the issue of reentrant phenomena in the surface transitions , raised in a previous paper , has been shortly discussed .
m. blume , phys .
rev . * 141 * , 517 ( 1966 ) .
capel , physica * 32 * , 966 ( 1966 ) .
d. mukamel and m. blume , phys .
rev . a * 10 * , 610 ( 1974 ) .
m. blume , v.j .
emery and r.b .
griffiths , phys .
rev . a * 4 * , 1071 ( 1971 ) .
a. benyoussef , n. boccara and m. saber , j. phys .
c * 19 * , 1983 ( 1986 ) .
a. benyoussef , n. boccara and m. el bouziani , phys .
b * 34 * , 7775 ( 1986 ) . l. peliti and s. leibler , j. phys .
( parigi ) lett . * 45 * , l-596 ( 1988 ) .
a. crisanti and l. peliti , j. phys .
a * 18 * , l543 ( 1985 ) .
x.p . jiang and m.r .
giri , j. phys .
c * 21 * , 995 ( 1988 ) .
i. tamura , j. phys .
jpn * 51 * , 3607 ( 1982 ) . c. buzano and a. pelizzola ,
physica a * 195 * , 197 ( 1993 ) . c. buzano and a. pelizzola , j. phys .
matt . * 5 * , 7319 ( 1993 ) .
k. binder , in _ phase transition and critical phenomena _ , vol
. 8 , edited by c. domb and j.l .
lebowitz ( academic , london , 1983 ) .
r. kikuchi , phys .
rev . * 81 * , 988 ( 1951 ) . g. an , j. stat .
* 52 * , 727 ( 1988 ) .
t. morita , j. stat .
phys . * 59 * , 819 ( 1990 ) .
a. pelizzola , phd thesis , politecnico di torino , 1995 ( in italian ) .
t. kaneyoshi , physica a * 163 * , 533 ( 1990 ) . c. buzano and a. pelizzola ,
physica a * 189 * , 333 ( 1992 ) .
r. kikuchi , j. chem .
* 60 * , 1071 ( 1974 ) .
r. kikuchi , j. chem .
phys . * 65 * , 4545 ( 1976 ) .
a. pelizzola , physica a * 211 * , 107 ( 1994 ) . d.p .
landau and k. binder , phys .
b * 41 * , 4633 ( 1990 ) .
a. pelizzola , j. magn .
magn . mat .
* 140144 * ( 1995 , in press ) . fig .
phase diagram topologies : heavy and light lines denote bulk and surface transitions , respectively , while solid and dashed lines denote second and first order transitions ; @xmath83 and @xmath82 denote critical and multicritical points . | the phase diagram of the blume capel model on a semi infinite simple cubic lattice with a ( 100 ) free surface is studied in the pair approximation of the cluster variation method .
six main topologies are found , of which two are new , due to the occurrence of a first order surface transition in the phase with ordered bulk , separating two phases with large and small surface order parameters .
the latter is a new phase and is studied in some detail , giving the behaviour of the order parameter profiles in two typical cases .
a comparison is made with the results of a low temperature expansion , where these are available , showing a great increase in accuracy with respect to the mean field approximation . |
magnetic vacuum birefringence or light - light interaction in vacuum at very low energies have yet to be observed .
several experimental efforts are underway@xcite@xcite to detect such effects .
indeed qed predicts non linear effects leading to birefringence and light - light scattering ( lbl ) through the box diagram.@xcite@xcite furthermore hypothetical ideas such as the existence of axion - like particles ( alps ) coupling to two photons@xcite@xcite or the existence of fermion / boson millicharged particles ( mcps)@xcite@xcite could generate both magnetic birefringence and magnetic dichroism .
finally the coupling of four photons through @xmath4 fluctuations is also possibile but an evaluation of such a contribution to the vacuum magnetic birefringence can not be extracted from indirect measurements.@xcite due to the quark masses , though , this last contribution can be expected to be very small .
these different contributions to the four photon interaction are summarized in figure [ 4photon ] . .
today the best limit on four photon interactions has been set by the pvlas collaboration@xcite with an upper bound on vacuum magnetic birefringence @xmath5 @xmath6 which , translated into light - light elastic scattering,@xcite@xcite results in an upper bound on the cross section @xmath7 @xmath8 the predicted qed value of the magnetic vacuum birefringence ( see below ) @xmath9 and light - light elastic scattering cross section @xmath10 are @xmath11 in the absence of matter , maxwell s equations can be obtained from the classical electromagnetic lagrangian density @xmath12 ( in s.i .
units ) @xmath13 it is well known that in this case the superposition principle holds thereby excluding light - light scattering and other non linear electromagnetic effects in vacuum . with the introduction of dirac s equation for electrons and heisenberg s uncertainty principle ,
euler and heisenberg in 1936@xcite derived a lagrangian density which leads to electromagnetic non linear effects even in vacuum .
for photon energies well below the electron mass and fields much smaller than their critical values , @xmath14 t , @xmath15 v / m , the euler - heisenberg lagrangian correction can be written as @xmath16\ ] ] where @xmath17 is the magnetic permeability of vacuum and @xmath18 with @xmath19 being the compton wavelength of the electron , @xmath20 the fine structure constant , @xmath21 the electron mass , @xmath22 the speed of light in vacuum .
this lagrangian correction allows four field interactions and can be represented , to first order , by the feynman diagrams shown in figure [ 4photon ] a ) and b ) .
figure [ 4photon ] a ) represents light by light scattering whereas figure [ 4photon ] b ) represents the interactions of real photons with a classical field leading to vacuum magnetic birefringence . to determine the magnetic birefringence of vacuum one can proceed by determining the electric displacement vector @xmath23 and magnetic intensity vector @xmath24 from the total lagrangian density @xmath25 by using the constitutive relations@xcite @xmath26 from these one obtains @xmath27 \label{d } \\
\vec h & = & \frac{\vec b}{\mu_{\rm 0 } } + \frac{a_{e}}{\mu_{\rm 0}}\big[4\big(\frac{e^{2}}{c^{2}}-b^{2}\big)\vec b - 14 \big(\frac{\vec e\cdot\vec b}{c^2}\big)\vec e\big ] \label{h } \end{aligned}\ ] ] with the @xmath23 and @xmath24 vectors one can use maxwell s equation in media to now describe light propagation in an external field .
it is evident that these will no longer be linear due to the non linear dependence of @xmath23 and @xmath24 with respect to @xmath28 and @xmath29 . by assuming a linearly polarized beam of light propagating perpendicularly to an external magnetic field @xmath30 there are two possible configurations : light polarization parallel to @xmath30 and light polarization perpendicular to @xmath30 . by substituting @xmath31 and @xmath32 in ( [ d ] ) and ( [ h ] )
one finds the following relations for the relative dielectric constants and magnetic permeabilities @xmath33 from these sets of equations two important consequences are apparent : the velocity of light in the presence of an external magnetic field is no longer @xmath22 and vacuum is birefringent with @xmath34 numerically this leads to the value given in equation ( [ deltan ] ) .
it is interesting to generalize the non linear electrodynamic lagrangian density correction by introducing three free parameters @xmath35 , @xmath36 and @xmath37 : @xmath38\ ] ] where @xmath39 whereas @xmath36 and @xmath37 are dimensionless parameters depending on the model .
with such a formulation the birefringence induced by an external magnetic field is @xmath40 this expression reduces to ( [ deltan ] ) if @xmath41 and @xmath42 .
it is thus apparent that @xmath43 depends only on @xmath44 whereas @xmath45 depends only on @xmath46 .
it is also noteworthy that if @xmath47 , as is the case in the born - infeld model , then there is no magnetically induced birefringence even though elastic scattering will be present.@xcite as mentioned above , two other important hypothetical effects could also cause @xmath48 in the presence of an external magnetic ( or electric ) field transverse to the light propagation direction .
these can be due either to neutral bosons weakly coupling to two photons called axion - like particles ( alp),@xcite@xcite or millicharged particles ( mcp).@xcite@xcite in this second case both fermions and spin-0 particles can be treated .
search for axions using laboratory optical techniques was experimentally pioneered by the bfrt collaboration@xcite and subsequently continued by the pvlas effort.@xcite initially , this second experiment published the detection of a dichroism induced by the magnetic field@xcite in vacuum .
such a result , although in contrast with the cast experiment,@xcite could have been due to axion - like particles . subsequently the result was excluded by the same collaboration@xcite after a series of upgrades to their apparatus and almost simultaneously the axion - like interpretation was excluded by two groups@xcite in a regeneration type measurement . however , the original publication revived interest in the optical effects which could be caused by alp s and later mcp s .
the lagrangian densities describing the interaction of either pseudoscalar fields @xmath49 or scalar fields @xmath50 with two photons can be expressed as ( for convenience , written in natural heavyside - lorentz units ) @xmath51 where @xmath52 and @xmath53 are the coupling constants .
therefore in the presence of an external uniform magnetic field @xmath54 a photon with electric field @xmath55 parallel to @xmath54 will interact with the pseudoscalar field whereas for electric fields perpendicular to @xmath54 no such interaction will exist .
for the scalar case the opposite is true : an interaction will exist if @xmath56 and will not if @xmath57 .
when an interaction is present , an oscillation between the photon and the pseudoscalar / scalar field will exist .
for photon energies above the mass @xmath58 of such particle candidates , a real production can follow .
this will cause an oscillation of those photons whose polarization allows an interaction into such particles . on the other hand ,
even if the photon energy is smaller than the particle mass , virtual production will follow and will cause a phase delay for those photons with an electric field direction allowing an interaction . the attenuation @xmath59 and phase delay @xmath60 for light with polarization allowing an interaction can be expressed , in both the scalar and pseudoscalar cases , as:@xcite @xmath61 where , in vacuum , @xmath62 , @xmath63 is the photon energy and @xmath64 is the magnetic field length .
the above expressions are in natural heavyside - lorentz units whereby 1 t @xmath65 ev@xmath66 and 1 m @xmath67 ev@xmath68 .
the phase delay @xmath60 is related to the index of refraction @xmath69 by @xmath70 therefore in the pseudoscalar case , where @xmath71 and @xmath72 , and in the scalar case , where @xmath73 and @xmath74 , the birefringence @xmath75 will be @xmath76 in the approximation for which @xmath77 ( small masses ) this expression becomes @xmath78 whereas for @xmath79 @xmath80 the different behavior of @xmath81 and @xmath82 with respect to @xmath64 in the two cases where @xmath77 and @xmath83 is interesting and leaves , in principle , a free experimental handle for distinguishing between various scenarios .
consider now the vacuum fluctuations of particles with charge @xmath84 and mass @xmath85 as discussed by gies and ringwald in references and .
the photons traversing a uniform magnetic field may interact with such fluctuations resulting in both a pair production if the photon energy @xmath86 and only a phase delay if @xmath87 .
furthermore , either fermions or spin-0 charged bosons could exist .
since we are discussing birefringence effects only the ( real ) index of refraction will be considered here .
- dirac fermions let us first consider the case in which the millicharged particles are dirac fermions ( df ) . as derived by tsai in reference the indices of refraction of photons with polarization respectively parallel and perpendicular to the external magnetic field
have two different mass regimes defined by a dimensionless parameter @xmath88 ( s.i .
units ) : @xmath89 it can be shown that@xcite @xmath90 with @xmath91 & \textrm { for } \chi \ll 1 \\ -\frac{9}{7}\frac{45}{2}\frac{\pi^{1/2}2^{1/3}\left(\gamma\left(\frac{2}{3}\right)\right)^{2}}{\gamma\left(\frac{1}{6}\right)}\chi^{-4/3}\left[\left(3\right)_{\parallel},\left(2\right)_{\perp}\right ] & \textrm { for } \chi\gg 1 \end{array}\right.\nonumber \label{idf}\ ] ] and @xmath92 in analogy to equation ( [ ae ] ) .
in the limit of large masses ( @xmath93 ) this expression reduces to ( [ index ] ) with the substitution of @xmath94 with @xmath95 and @xmath85 with @xmath96 in equation ( [ ndf ] ) .
the dependence on @xmath97 remains the same as for the well known qed prediction . for small masses ( @xmath98 )
the index of refraction now also depends on the parameter @xmath99 resulting in a net dependence of @xmath69 with @xmath100 rather than @xmath101 . in both mass regimes , a birefringence
is induced : @xmath102 a_{\epsilon } b_{\rm ext}^{2}=\\ & = & \left\{\begin{array}{ll } 3 a_{\epsilon } b_{\rm ext}^{2 } & \textrm { for } \chi \ll 1 \\ -\frac{9}{7}\frac{45}{2}\frac{\pi^{1/2}2^{1/3}\left(\gamma\left(\frac{2}{3}\right)\right)^{2}}{\gamma\left(\frac{1}{6}\right)}\chi^{-4/3}a_{\epsilon } b_{\rm ext}^{2 } & \textrm { for } \chi\gg 1 \end{array}\right.\nonumber \label{deltandf}\end{aligned}\ ] ] - spin-0 charged bosons very similar expressions to the dirac fermion case can also be obtained for the spin-0 ( s0 ) charged particle case.@xcite again there are two mass regimes defined by the same parameter @xmath88 of expression ( [ chi ] ) . in this case
the indices of refraction for the two polarization states with respect to the magnetic field direction are @xmath103 with @xmath104 & \textrm { for } \chi \ll 1 \\ -\frac{9}{14}\frac{45}{2}\frac{\pi^{1/2}2^{1/3}\left(\gamma\left(\frac{2}{3}\right)\right)^{2}}{\gamma\left(\frac{1}{6}\right)}\chi^{-4/3}\left[\left(\frac{1}{2}\right)_{\parallel},\left(\frac{3}{2}\right)_{\perp}\right ] & \textrm { for } \chi\gg 1 \end{array}\right.\nonumber \label{is0}\ ] ] the vacuum magnetic birefringence is therefore @xmath105 a_{\epsilon } b_{\rm ext}^{2}=\\ & = & \left\{\begin{array}{ll } -\frac{6}{4 } a_{\epsilon } b_{\rm ext}^{2 } & \textrm { for } \chi \ll 1 \\ \frac{9}{14}\frac{45}{2}\frac{\pi^{1/2}2^{1/3}\left(\gamma\left(\frac{2}{3}\right)\right)^{2}}{\gamma\left(\frac{1}{6}\right)}\chi^{-4/3}a_{\epsilon } b_{\rm ext}^{2 } & \textrm { for } \chi\gg 1 \end{array}\right.\nonumber \label{deltans0}\end{aligned}\ ] ] as can be seen there is a sign difference in the birefringence @xmath106 induced by an external magnetic field in the presence of dirac fermions with respect to the case in which spin-0 particles exist
. figures [ 4photon ] c ) and d ) show the feynman diagrams for the @xmath107 contribution to the vacuum magnetic birefringence .
the effective lagrangian density for this correction has been evaluated by different authors@xcite@xcite and can be expressed as @xmath108\ ] ] this lagrangian leads to an extra correction @xmath109 of the vacuum magnetic birefringence given in equation ( [ birifqed ] ) @xmath110 resulting in a 1.45 % correction .
the aim of the pvlas collaboration is to build an apparatus capable of measuring very small ellipticities and rotations . in particular
the ultimate goal is to measure the vacuum magnetic birefringence predicted by the euler - heisenberg lagrangian . at present
the best experimental result is an upper bound on @xmath1.@xcite this bound is a factor of about 5000 above the predicted value . a sensitive ellipsometer designed to measure small birefringences
can also be used to measure small rotations due to dichroism .
although qed does not predict dichroism ( photon splitting is unmeasurably small ) , alp s and mcp s could .
as will be shown below a birefringence @xmath106 will induce an ellipticity @xmath111 on a linearly polarized beam of light given by @xmath112 where @xmath113 is the effective path length within the birefringent region with birefringence @xmath106 and @xmath114 is the wavelength of the light traversing it .
the induced ellipticity also depends on the angle @xmath115 between the light polarization and the magnetic field direction . in the qed case @xmath106
depends quadratically on the magnetic field @xmath116 .
therefore the magnetic field region must be as long as possible , the magnetic field as intense as possible and the wavelength small .
finally the expected ellipticity must be compared to the different noise sources present and to the maximum available integration time . experimentally @xmath113 can be made very long by using a very high finesse fabry - perot cavity .
in fact given a birefringent region of length @xmath64 within a fabry - perot cavity of finesse @xmath117 the effective path length is @xmath118 . today
finesses @xmath119 can be obtained .
high magnetic fields can be obtained with superconducting magnets but as we will see below it is desirable to have a time dependent field either by ramping it , thereby changing @xmath106 or by rotating the field direction , thereby changing @xmath115 .
this makes superconducting magnets far less appealing than permanent magnets which , today , can reach fields above 2.5 t. furthermore permanent magnets are relatively inexpensive to buy , have no running costs and have 100% duty cycle allowing in principle very long integration times .
as for the wavelength we are working with a nd : yag laser emitting radiation at @xmath120 nm .
frequency doubled versions exist and could double the induced ellipticity but at the moment the highest finesses have been obtained without the frequency doubling .
lastly it is necessary to make the magnetic field time dependent to move away from dc measurements and limit @xmath121 noise .
two detection schemes exist : homodyne detection or heterodyne detection .
this second technique has been adopted in the pvlas collaboration .
a scheme of the ellipsometer is shown in figure [ scheme ] .
the input polarizer linearly polarizes the laser beam of intensity @xmath122 which then enters the sensitive region delimited by the fabry - perot cavity mirrors .
the laser is phase locked to this cavity thus increasing the optical path length within the magnetic field by a factor @xmath123 where @xmath124 is the finesse of the cavity .
after the cavity the laser beam passes through a photo - elastic ellipticity modulator ( pem ) which adds a known time dependent ellipticity @xmath125 to the beam .
this modulator ellipticity adds to the ellipticity @xmath126 acquired within the magnetic field region . after the pem the beam passes through the analyzer which selects the polarization perpendicular to the input polarization .
a photodiode detects @xmath127 and its fourier spectrum is then analyzed .
to better understand what follows it is useful to present some numerical values of the different quantities involved in the pvlas experiment .
considering the vacuum magnetic birefringence due to the euler - heisenberg lagrangian , let us determine the ellipticity we expect in the apparatus under construction .
we will have a total magnetic field length @xmath128 m with a field intensity @xmath129 t resulting in @xmath130 t@xmath66 m .
at present we are running with a finesse @xmath131 . in the past
we have reached a maximum finesse value of @xmath132 .
such values have also been published by other authors .
we can therefore assume a value of @xmath133 for our calculation .
putting these numbers together leads to @xmath134 assuming a maximum integration time @xmath135 s and a signal to noise ratio snr = 1 implies that the sensitivity must be @xmath136 as discussed in ref .
the ultimate shot noise ellipticity sensitivity limit with the heterodyne technique depends only on the current generated in the photodiode . assuming the power output from the cavity @xmath137 mw and the quantum efficiency of the diode @xmath138 a / w . @xmath139 at present our sensitivity with @xmath137 mw is about @xmath140 at the frequency of interest .
work is underway to try to understand the noise present .
considering the coherence of the light source a full treatment of the system can be done with the jones matrix formalism.@xcite for the purpose of our discussion let the laser beam propagate along the @xmath141 axis and let the incoming ( linear ) polarization define the @xmath142 axis ( figure [ sistrif ] ) .
the jones matrix for a uniaxial birefringent element of length @xmath64 is given by @xmath143 where @xmath144 ( @xmath145 ) is the induced ellipticity acquired by the light , @xmath115 represents the angle between the slow axis ( @xmath146 ) of the medium and the @xmath142 axis .
furthermore @xmath147 with @xmath148 the phase delay between the parallel and perpendicular polarization components acquired in the length @xmath64 .
the entrance polarizer defines the input electric field @xmath149 which , after the magnetic field region , will be @xmath150 assuming no losses , the power @xmath127 after the analyzer ( polarizer crossed with respect to the entrance polarizer ) will therefore be @xmath151 the output power is proportional to @xmath152 and given the predicted value results in an unmeasurably small intensity component . by adding a known sinusoidal ellipticity @xmath153 generated with the pem ,
the ellipticity signal @xmath144 is linearized .
in fact the jones matrix for the modulator is the same as @xmath154 with @xmath115 set at an angle of @xmath155 ( @xmath156 ) : @xmath157 the resulting vector describing the electric field after the modulator will be @xmath158 neglecting second order terms , the power @xmath127 after the analyzer will be @xmath159\ ] ] which now depends linearly on the ellipticity @xmath144 . to complete the discussion
, one finds experimentally that static and slowly varying ellipticities , indicated as @xmath160 , are always present in an real apparatus and that two crossed polarizers have an intrinsic extinction ratio @xmath161 , mainly due to imperfections in the calcite crystals .
furthermore , losses in the system reduce the total light reaching the analyzer which we will now indicate as @xmath162 . therefore , taking into account an additional spurious ellipticity term @xmath160 ( since @xmath163 these terms commute and therefore add up algebraically ) and a term proportional to @xmath161 , the total power at the output of the analyzer will be @xmath164\simeq\\ & \simeq&i_{\rm out}\left[\sigma^2+\eta(t)^2+\alpha(t)^{2}+2\eta(t)\psi\sin2\vartheta+2\eta(t)\alpha(t)\right]\end{aligned}\ ] ] to be able to distinguish the large term @xmath165 from the term @xmath166 , @xmath167 is also modulated in time .
this can be done by either ramping the magnetic field intensity ( varying therefore @xmath144 ) or by rotating the magnetic field direction ( varying @xmath115 ) .
the final expression , explicitly indicating the time dependence of @xmath144 and @xmath115 , for the power at the output of the analyzer is therefore @xmath168 \label{spurious}\ ] ] to increase the ellipticity induced by the birefringent region of length @xmath64 one can increase the number of passes through it . either a multi - pass cavity or a fabry - perot cavity can be used for this purpose . in the pvlas experiment described below , a fabry - perot has been chosen . in a multi - pass cavity
the induced ellipticity is proportional to the number of passes @xmath169 through the region . with a fabry - perot cavity
the analogy to a multi - pass cavity is not immediate since one is dealing with a standing wave .
let @xmath170 , @xmath171 be the transmission and reflection coefficients , and @xmath172 the losses of the mirrors of the cavity such that @xmath173 .
let @xmath174 be the length of the cavity and @xmath175 the roundtrip phase for a beam of wavelength @xmath114 .
then the jones matrix for the elements of the ellipsometer after the entrance polarizer is @xmath176^{n}}\cdot\mathbf{bf}\ ] ] where @xmath177 is the analyzer jones matrix and @xmath178 describes the spurious ellipticity mainly due to the mirrors of the cavity itself . because @xmath179 , @xmath180 can be rewritten as @xmath181^{-1}}\cdot\mathbf{bf } \label{fp}\ ] ] with @xmath182 the identity matrix . with the laser phase locked to the cavity so that @xmath183 , where @xmath184 is an integer number , the electric field at the output of the system will be @xmath185 and
the power , including losses , @xmath186 this expression is at the basis of the ellipsometer in the pvlas apparatus.@xcite small ellipticities add up algebraically and the fabry - perot multiplies the single pass ellipticity @xmath187 , generated within the cavity , by a factor @xmath188 , where @xmath124 is the finesse of the cavity .
the ellipticity signal to be detected is therefore @xmath189 . in the pvlas experiment , @xmath190 and the magnetic field direction
is rotated at an angular velocity @xmath191 .
a fourier analysis of the power @xmath192 of equation ( [ iout ] ) results in four main frequency components each with a definite amplitude and phase .
these are reported in table [ components ] .
c|c|c|c frequency & fourier component & intensity/@xmath122 & phase + @xmath193 & @xmath194 & @xmath195 & @xmath196 + @xmath197 & @xmath198 & @xmath199 & @xmath200 + @xmath201 & @xmath202 & @xmath203 & @xmath204 + @xmath205 & @xmath206 & @xmath207 & @xmath208 + [ components ] the presence of a component at @xmath201 in the signal identifies an induced ellipticity within the fabry - perot cavity .
furthermore the phase of this component must satisfy the value in table [ components ] .
although the previous pvlas apparatus@xcite setup at lnl in legnaro , italy , set best limits on magnetic vacuum birefringence and photon - photon elastic scattering at low energies several limitations were present in this apparatus : * stray field due to the superconducting magnet when operating at high fields * limited running time due to liquid helium consumption in the rotating cryostat * high running costs * seismic noise * with a single magnet a zero measurement with the same experimental conditions as with the field on is not possible to solve the first two points in the new setup we chose to work with permanent magnets instead of the superconducting magnet . in this way
the stray field will be much smaller and the duty cycle will be 100% .
furthermore studies on the pvlas - lnl setup revealed limitations due to seismic noise . due to the size and configuration of the optical benches this problem
could not be solved directly on site . as was shown in it
is necessary to seismically isolate the whole ellipsometer apparatus on a single optical bench . for these reasons
a complete new setup is being rebuilt on a single granite optical bench .
the last point in the above list deserves some attention . a zero measurement in an experimental condition as close to the signal configuration as possible is absolutely mandatory after the experience with the pvlas - lnl apparatus .
especially when dealing with such a sensitive apparatus . therefore instead of using a single dipole magnet , two magnets will be used . by orienting the fields of the two magnets at @xmath209 and assuming the magnets to be identical the net ellipticity generated by the magnetic birefringence is zero . running the system with the magnets parallel and perpendicular
will allow the identification of a real physical signal with respect to some spurious signal due to stray field .
a verification of this idea was done with our test setup in ferrara using the cotton - mouton effect of oxygen .
the principle of the test ellipsometer is identical to the final system in construction .
a photograph of the test apparatus is shown in figure [ testsetup ] . at the center
one can see the two permanent magnets each generating a 20 cm long magnetic field of intensity @xmath210 t. the whole optical setup is placed on a seismically isolated optical bench whereas the magnets are supported by a structure placed on the floor thereby mechanically isolated from the optics . .
figure [ cm_o2 ] shows the fourier spectrum around the carrier frequency @xmath197 .
clear sidebands can be seen at twice the rotation frequency of the magnets ( @xmath211 hz ) . shown in grey
is the fourier transform with the magnets in a parallel configuration whereas in black the magnets are perpendicular .
the signal attenuation factor in the perpendicular configuration with respect to the parallel one is about 80 .
although we also see sidebands at once the magnet rotation frequency ( which should in principle not be there ) which means we may have a small spurious component even a @xmath212 , we can conclude that the parameter @xmath213 for the two magnets is equal to within about 1 - 2% and most importantly that the principle is correct . .
we believe this improvement will be crucial in understanding the ellipsometer and that a measurement with two magnets whose field directions can be changed is imperative .
there is no way of guaranteeing the authenticity of a signal observed with only one magnet or with several magnets all in a parallel configuration . with the ferrara test setup ( figure [ testsetup ] ) measurements
have been performed to understand its limits and optimize the new apparatus in construction .
two different noise sources exist and are under study : wideband noise and signals at the magnet rotation frequency and its harmonics .
below we briefly report some results in the two cases .
measurements were first performed without the fabry - perot cavity .
we successfully exclude noise sources from readout electronics and optical elements other than the cavity mirrors reaching the expected sensitivity of @xmath214 and a noise floor of @xmath215 with @xmath216 s integration time .
with the introduction of the cavity with finesse @xmath217 the noise increased to @xmath218 at about 6 hz .
this was significantly more than what was expected from the reduction of @xmath162 due to cavity losses .
this unexplained noise is under study and we suspect variations of the intrinsic birefringence of the mirrors .
we also showed that the magnet rotation did not contribute to the wideband noise indicating a good isolation between the magnet support and optical setup . with the magnets in rotation we often observe ellipticity peaks varying from a few @xmath219 to a few @xmath220 whereas sometimes such peaks are not present .
the frequencies of these peaks are at harmonics of the magnet rotation frequency .
the variability of these peaks from one run to another seems to depend ( in a non reproducible way ) on the adjustment of the input and output polarizers which are done with motorized stages . to study the dependence of such peaks on the magnet orientation ,
a field probe is present near the output side magnet .
changing the relative orientation of the two magnets does not change the amplitude of these peaks but does change their phase indicating that the more sensitive part of the apparatus seems to be the entrance optics .
all the motorized stages have small electric motors which couple to the rotating magnetic field and may introduce beam jitter and therefore ellipticity .
the substitution of all these stages is programmed shortly . . with the apparatus in a condition in which the peak a @xmath212 is not present measurements of a few hours
have been done .
the magnet rotation frequency was @xmath221 hz . in figure [ limite ] we report the histogram of the ellipticity noise in a narrow band around @xmath212 : @xmath222 hz .
the integration time in this example was @xmath223 s. the probability density function for a noise amplitude with equal standard deviations @xmath224 for the in phase and quadrature components is the rayleigh function : @xmath225 . in figure [ limite ]
we have superimposed a fit with the rayleigh function in which the ellipticity standard deviation is @xmath226 .
a vertical line in the same figure indicates the value in the fourier spectrum bin corresponding to exactly @xmath227 hz .
given the finesse @xmath217 , @xmath228 m , @xmath229 t and @xmath230 nm the value @xmath226 translates , at 95 % c.l . , in a birefringence limit induced by the magnetic field of @xmath231 the parameter @xmath1 and the light - light elastic scattering cross section limit can also be deduced ( at 95 % c.l . ) : @xmath232 although not always reproducible , these limits are about a factor 2 better than the best previously published limits obtained with the pvlas - lnl apparatus.@xcite
we have presented the physics the pvlas experiment is aiming at studying and have briefly discussed the experimental method .
noise sources are being studied on a bench - top small apparatus in ferrara , italy , in view of the construction of the final apparatus with which we hope to measure for the first time the magnetic birefringence of vacuum due to vacuum fluctuations .
we have discussed the importance of using two dipole magnets instead of only one whose directions can be made perpendicular to each other in order to have a zero effect condition with the magnetic field present .
this is vital to study and eliminate spurious signals generated by the rotating field .
a proof of principle measurement using the cotton - mouton effect in oxygen gas was done with the test apparatus in ferrara , italy .
finally noise floor measurements were performed in conditions in which spurious peaks were not present .
a new limit on the parameter @xmath1 describing non linear electrodynamic effects in vacuum has been obtained : @xmath2 t@xmath3 .
this value improves the previous one by a factor 2.2 .
00 d. bakalov _ et al .
_ , hyperfine interactions * 114 * ( 1998 ) 103
. + d. bakalov _ et al .
_ , quantum semicl . opt . * 10 * ( 1998 ) 239 .
e. zavattini _ et al .
d * 77 * ( 2008 ) 032006 .
e. lundstrm _
et al . _ ,
* 96 * ( 2006 ) 083602 .
d. tommasini _ et al .
a * 77 * ( 2008 ) 042101 .
a. n. luiten and j. c. petersen , phys .
lett . a * 330 * ( 2004 ) 429
. a. n. luiten and j. c. petersen , phys .
a * 70 * ( 2004 ) 033801 .
r. battesti _ et al .
j. d * 46 * , ( 2008 ) 323 .
ni , chinese j. phys .
* 34 * ( 1996 ) 962 .
p. pugnat _
et al . _ ,
a * 56 * ( 2006 ) c193 .
t. heinzl _
et al . _ , optics comm . * 267 * ( 2006 ) 318 .
h. euler and b. kochel , naturwiss .
* 23 * ( 1935 ) 246 .
+ w. heisenberg and h. euler , z. phys .
* 98 * ( 1936 ) 718 .
+ v. s. weisskopf , kgl .
danske vid .
, math .- fys . medd .
* 14 * ( 1936 ) 6 .
+ j. schwinger , phys . rev .
* 82 * ( 1951 ) 664 .
r. baier and p. breitenlohner , acta phys .
austriaca * 25 * ( 1967 ) 212 .
+ r. baier and p. breitenlohner , nuovo cimento * 47 * ( 1967 ) 261 .
+ s. l. adler , ann .
* 67 * ( 1971 ) 559 .
+ z. bialynicka - birula and i. bialynicki - birula , phys .
d * 2 * ( 1970 ) , 2341 .
e. iacopini and e. zavattini , phys .
b * 85 * ( 1979 ) 151 .
b. de tollis , nuovo cimento * 35 * ( 1965 ) 1182 . + b. de tollis , nuovo cimento * 32 * , ( 1964 ) 757 .
r. karplus _ et al .
_ , phys . rev .
* 83 * ( 1951 ) 776 .
d. a. dicus _
d * 57 * ( 1998 ) 2443 .
d. bernard _
et al . _ ,
j. d * 10 * ( 2000 ) 141 .
f. moulin _
_ , z. phys .
c * 72 * ( 1996 ) 607 .
j. hassinski _
* 74 * ( 2006 ) 678 - 681 .
v. i. denisov _
et al . _ ,
d * 69 * ( 2004 ) 066008 .
l. maiani _
lett . b*175 * ( 1986 ) 359 .
p. sikivie , phys .
* 51 * ( 1983 ) 1415 .
m. gasperini , phys .
lett . * 59 * ( 1987 ) 396 .
g. raffelt and l. stodolsky , phys .
d * 37 * ( 1998 ) 1237 .
m. ahlers _ et al .
d * 75 * ( 2007 ) 035011 .
h. gies _ et al .
* 97 * ( 2006 ) 140402 .
w. y. tsai and t. erber , phys .
d * 12 * ( 1975 ) 1132 .
j. k. daugherty _
et al . _ ,
j. * 273 * ( 1983 ) 761 . c. schubert , nucl .
b585 * ( 2000 ) 407 .
m. davier et al . , arxiv:0908.4300[hep - ph ] and refs .
therein a. nyffeler , arxiv:1001.3970[hep - ph ] and refs .
therein m. bregant _ et al .
_ , phys rev .
d * 78 * ( 2008 ) 032006 .
f. della valle _
et al . _ , optics comm . * 293 * ( 2010 ) 4194 .
r. cameron _
et al . _ ,
d * 47 * ( 1993 ) 3707 .
e. zavattini _ et al .
* 96 * ( 2006 ) 110406 .
k. zioutas _ et al .
lett . * 94 * ( 2005 ) 121301 . c. robilliard _
et al . _ , phys .
* 99 * ( 2007 ) 190403 .
m. fouch _
_ , phys . rev .
d * 78 * ( 2008 ) 032013 .
a. s. chou _
et al . _ , arxiv:0710.3783 + a. s. chou _
et al . _ ,
* 100 * ( 2008 ) 080402 . v. i. ritus , zh .
fiz * 69 * ( 1975 ) 1517 [ sov
jetp * 42 * ( 1975 ) 774 ] .
d. bakalov , infn / ae-94/27 ( 1994 ) sis publication lnf g. dunne , arxiv : hep - th/0406216v1 r. c. jones , j. opt .
soc . am . 38 ( 1948 ) 671 .
+ e. hecht , _ optics _ , 2nd ed .
, addison - wesley , san francisco ( 1987 ) . g. zavattini _
et al . _ , applied physics b * 83 * ( 2006 ) 571 . | we describe the principle and the status of the pvlas experiment which is presently running at the infn section of ferrara , italy , to detect the magnetic birefringence of vacuum .
this is related to the qed vacuum structure and can be detected by measuring the ellipticity acquired by a linearly polarized light beam propagating through a strong magnetic field .
such an effect is predicted by the euler - heisenberg lagrangian .
the method is also sensitive to other hypothetical physical effects such as axion - like particles and in general to any fermion / boson millicharged particle .
here we report on the construction of our apparatus based on a high finesse ( @xmath0 ) fabry - perot cavity and two 0.9 m long 2.5 t permanent dipole rotating magnets , and on the measurements performed on a scaled down test setup . with the test setup
we have improved by about a factor 2 the limit on the parameter @xmath1 describing non linear electrodynamic effects in vacuum : @xmath2 t@xmath3 @ 95% c.l . |
the australian square kilometre array pathfinder ( askap ) is currently being built at the proposed site for the square kilometer array ( ska ) in western australia .
askap will be a 36-element array of 12-m antennas with baselines of up to several km ; it is well - described in @xcite .
it will use checkerboard - style phased - array feeds @xcite with dithering to achieve roughly uniform sensitivity over a field of view ( fov ) of 30 square degrees .
several primary beams will be used to cover the full fov , with the number of beams increasing with frequency so as to maintain the same size fov at all frequencies between 700mhz and 1800mhz .
roughly 300mhz of bandwidth will be available at any time .
askap will also offer multiple tied - array beams , steerable within the fov , for projects interested in point - source work .
the projected system temperature is roughly 50k .
the goals of the askap project are to demonstrate technologies ( specifically inexpensive parabolic reflectors and phased - array feeds ) relevant to the ska , to carry out science related to the ska key science projects @xcite , to establish a radio astronomy site in western australia and to build a user base for the ska @xcite .
the choice has been made to optimize the combination of antenna layout , fov , observing frequencies and system temperature to allow rapid surveys of neutral hydrogen in our galaxy and nearby galaxies , along with continuum surveys of galaxies to high redshift .
polarization surveys and exploration of transient phenomena are also priorities .
the full science case for askap can be found in @xcite .
askap will reduce all imaging data taken with one or more real - time pipelines , archiving images and spectral lines as required .
observers will retrieve data from the archive rather than operating the telescope themselves .
all archived data will be publically available @xcite . to ensure that good analysis pipelines are developed , observers have been encouraged to organize into `` science survey '' teams , 10 of which
have been chosen to carry out `` design studies '' in the period 2009 - 2011 .
advantages of joining a science survey project ( ssp ) include ongoing communication with and support from cass during askap development , ensuring that ssp scientists will be well - placed to understand and extract science from the data the moment they appear in the archive . with a total collecting area similar to that of the parkes telescope , but
significantly higher system temperature , askap is perhaps not the most obvious choice for pulsar observations .
nevertheless , pulsars are at the core of one of the 5 ska key science projects ( `` strong - field tests of gravity using pulsars and black holes '' ) and form one of the two driving projects for the phase i ska . since pulsar observations have historically been carried out primarily using large single dishes , we have formed the coast ( `` compact objects with askap : surveys and timing '' ) ssp collaboration to work on the technical aspects of migrating pulsar observations to interferometers while carrying out niche pulsar science in the areas where askap can excel .
the coast ssp proposal comprises both search and timing observations , aimed at a broad range of pulsar - related science . despite somewhat lower point - source sensitivity than parkes ( nominally about 65% , but it could be about 90% if the system temperature can be reduced to 35k ) , askap can still achieve good signal - to - noise ratios on millisecond pulsars ( msps ) with sufficiently long integrations .
it is in a low - interference environment and should be somewhat more robust to interference than a single dish . these properties could allow it to take on part of the southern - hemisphere burden of monitoring msps for the purposes of contraints on a background of gravitational waves , and perhaps participate in an eventual detection ( e.g. , * ? ? ?
this would make askap part of the international pulsar timing array project .
while the msps are being monitored , the rest of the 30-square - degree askap fov need not go to waste ; there are often other pulsars visible , and in the galactic plane there are regions where 20 or more pulsars are accessible at once ( though not all of these contain msps ) . with new pulsars being discovered in ongoing surveys ( e.g. , * ? ? ?
* ) , the surface density of sources will only increase .
in general , we expect most of the `` secondary '' pulsars in the fov to be young .
many may glitch or show large amounts of timing systematics ( e.g. , * ? ? ? * ) and/or be targets for high - energy observations , for example with the fermi gamma - ray observatory .
both young and millisecond pulsars are exciting targets for gravitational - wave observatories such as ligo . there are therefore multiple motivations for ongoing monitoring to produce accurate radio ephemerides , and to record time - series data for pulsars known to glitch .
the myriad science areas accessible to pulsar observations will all benefit from the discovery of previously unknown pulsars , either by finding unique objects and test - cases , or by the improved statistics of a larger and more comprehensive sample .
again , it is necessary to iron out the technical details of how to accomplish pulsar surveys with an interferometer , in order to fulfill part of one of the key science projects of the ska @xcite .
one technique aimed at discovering primarily millisecond and fast binary pulsars will be to search pulsar - like point sources found in the large continuum surveys such as emu ( evolving map of the universe ) and the transient surveys such as vast ( variability and slow transients ) and craft ( commensal real - time askap fast transients ) that will be carried out with askap .
emu will find weak point sources , which we can examine for steep spectral indices and/or strong polarization , which may be signposts to pulsars .
vast and craft will find stronger sources that are time- and/or frequency - variable ; the scintillation properties may also indicate that pulsars are present .
work is underway to examine current point - source catalogs to determine the optimal criteria to use . using tied - array beams on the point sources
, we will be able to record fast - sampled , high - frequency - resolution data and thus be sensitive to millisecond pulsars .
the selection effects will be different from those of traditional wide - area searches , possibly opening up an entirely new pulsar discovery space . once again aiming to develop techniques that will be needed for the ska
, we also plan to carry out a search of a wide area of sky , using the data recorded from the entire askap fov . given the data rate limitations from the correlator , such a search will necessarily be restricted to comparatively wide frequency channels of a few mhz and to sample rates of milliseconds ; even then the development of the necessary correlator modes will be some years off .
the data reduction challenge is also significant , given the need to search either every pixel in the fov or else the uv plane itself .
we may choose to use only the inner core of 30 antennas to restrict the baselines to 2 km and reduce the number of pixels .
this search will of necessity be sensitive primarily to slower pulsars with spin periods greater than @xmath0 milliseconds , but there are a number of exciting types of objects in this category , including young pulsars , radio magnetars and even double - neutron - star binaries . to maximize the scientific output from such a survey
, we intend to conduct searches in the lowest part of the askap frequency band ( 7001000mhz , a regime not considered so far in wide - area searches ) and a target zone just off the galactic plane to minimize dispersion and scattering , while expecting both mildly recycled pulsars and young pulsars to be in our fields .
another likely target for such a survey will be the magellanic clouds , which we can cover deeply in just one pointing each .
through the design study , we are investigating possible solutions to several algorithmic and hardware problems as well as planning trial observations with the beta ( boolardy engineering test array ) prototype of 6 antennas that should be available in 2011 - 2 .
although the timing science plan is well - understood , one area that is being investigated is the optimal selection of particular fields on the sky that will place the most interesting pulsars near the centre of the fov to minimize polarization impurities while still maintaining a good number of interesting secondary pulsars . since we expect the list of primary pulsars to evolve as more objects are discovered , we need a reliable algorithm for fov selection rather than a rigid list .
the point - source - search selection criteria are under active exploration using existing catalogs from the atlas survey @xcite .
population syntheses for both the point - source and wide - area surveys are being carried out with two different code packages ; see the article by bouchard in this volume for details
. meanwhile , algorithm development for the wide - area searches is proceeding .
currently this includes simulation of both point - like and pulsed sources on the sky and software to detect pulsations along individual baselines .
the next steps include searches of the simulated pixelized sky and development of an algorithm to indentify pulsar signals from unknown sky positions using multiple baselines simultaneously , effectively searching directly in the uv plane .
the success of the intermediate stages of this software are shown in figure [ fig : source ] , which demonstrates the successful recovery of simulated point sources .
we will need to provide pulsar backend hardware to acquire the signals and process them .
this is made significantly easier relative to what is required at most telescopes by the askap engineers intention to provide us with tied - array beam signals that are already nyquist - sampled and separated into few - mhz channels , or else in a full - stokes digitized filterbank format .
therefore most of our data acquisition needs can be satisfied by computing power to handle the coherent dedispersion task for msp timing , to fold or make dedispersed timeseries data for the slower pulsars and simply to record the filterbank data for the point - source searches .
disk space will be the main requirement for the last type of data , while small ( possibly gpu - based ) clusters should be able to handle the computational tasks .
the wide - area survey will require significantly more and faster disks ; this problem will be addressed closer to the time when these observations might be scheduled . our observations with the 6-element beta telescope will focus on timing several bright pulsars simultaneously . in particular , we intend to observe two pulsars separated across the field of view to characterize the stability of the telescope response and tied - array beams in particular over data spans of several hours .
the murchison radio - astronomy observatory ( mro ) is jointly funded by the commonwealth government of australia and state government of western australia and managed by the csiro .
we acknowledge the wajarri yamatji people as the traditional owners of the observatory site .
pulsar and askap research at ubc is supported by nserc discovery and sro grants .
d. r. deboer , r. g. gough , j. d. bunton , t. j. cornwell , r. j. beresford , s. johnston , i. j. feain , a. e. schinckel , c. a. jackson , m. j. kesteven , a. chippendale , g. a. hampson , j. d. osullivan , s. g. hay , c. e. jacka , t. w. sweetnam , m. c. storey , l. ball , and b. j. boyle , _ ieee proceedings _ * 97 * , 15071521 ( 2009 ) .
s. johnston , m. bailes , n. bartel , c. baugh , m. bietenholz , c. blake , r. braun , j. brown , s. chatterjee , j. darling , a. deller , r. dodson , p. g. edwards , r. ekers , s. ellingsen , i. feain , b. m. gaensler , m. haverkorn , g. hobbs , a. hopkins , c. jackson , c. james , g. joncas , v. kaspi , v. kilborn , b. koribalski , r. kothes , t. l. landecker , e. lenc , j. lovell , j. macquart , r. manchester , d. matthews , n. m. mcclure - griffiths , r. norris , u. pen , c. phillips , c. power , r. protheroe , e. sadler , b. schmidt , i. stairs , l. staveley - smith , j. stil , r. taylor , s. tingay , a. tzioumis , m. walker , j. wall , and m. wolleben , _ proc .
* 24 * , 174188 ( 2007 ) .
s. johnston , r. taylor , m. bailes , n. bartel , c. baugh , m. bietenholz , c. blake , r. braun , j. brown , s. chatterjee , j. darling , a. deller , r. dodson , p. edwards , r. ekers , s. ellingsen , i. feain , b. gaensler , m. haverkorn , g. hobbs , a. hopkins , c. jackson , c. james , g. joncas , v. kaspi , v. kilborn , b. koribalski , r. kothes , t. landecker , a. lenc , j. lovell , j. macquart , r. manchester , d. matthews , n. mcclure - griffiths , r. norris , u. pen , c. phillips , c. power , r. protheroe , e. sadler , b. schmidt , i. stairs , l. staveley - smith , j. stil , s. tingay , a. tzioumis , m. walker , j. wall , and m. wolleben , _ experimental astronomy _ * 22 * , 151273 ( 2008 ) .
m. j. keith , a. jameson , w. van straten , m. bailes , s. johnston , m. kramer , a. possenti , s. d. bates , n. d. r. bhat , m. burgay , s. burke - spolaor , n. damico , l. levin , p. l. mcmahon , s. milia , and b. w. stappers , _ mnras _ * 409 * , 619627 ( 2010 ) .
r. p. norris , j. afonso , p. n. appleton , b. j. boyle , p. ciliegi , s. m. croom , m. t. huynh , c. a. jackson , a. m. koekemoer , c. j. lonsdale , e. middelberg , b. mobasher , s. j. oliver , m. polletta , b. d. siana , i. smail , and m. a. voronkov , _ aj _ * 132 * , 24092423 ( 2006 ) . | the australian square kilometre array pathfinder ( askap ) is a 36-element array with a 30-square - degree field of view being built at the proposed ska site in western australia .
we are conducting a design study for pulsar observations with askap , planning both timing and search observations .
we provide an overview of the askap telescope and an update on pulsar - related progress .
address = dept . of physics and astronomy , ubc , 6224 agricultural road , vancouver ,
bc v6 t 1z1 canada address = csiro astronomy and space science , australia telescope national facility , csiro , po box 76 , epping , nsw 1710 , australia address = cresst / usra , nasa goddard space flight center , code 662 , greenbelt , md 20771 , usa address = max - planck - institut fr extraterrestrische physik , giessenbachstrasse 1 , 85748 garching , germany address = dept . of physics and astronomy , ubc , 6224 agricultural road , vancouver ,
bc v6 t 1z1 canada address = department of physics , mcgill university , 3600 university street , montreal , quebec , h3a 2t8 , canada address = swinburne university of technology , centre for astrophysics and supercomputing , mail h39 , po box 218 , vic 3122 , australia address = inaf - osservatorio astronomico di cagliari , localit poggio dei pini , strada 54 , i-09012 capoterra , italy address = max planck institut fr radioastronomie , auf dem hgel 69 , 53121 bonn , germany address = astronomy department , cornell university , ithaca , ny 14853 , usa address = icrar / curtin institute of radio astronomy , gpo box u1987 , perth , wa 6845 australia address = astronomy department , cornell university , ithaca , ny 14853 , usa address = department of physics and astronomy , franklin and marshall college , p.o .
box 3003 , lancaster , pa 17604 , usa address = icrar / university of western australia , fairway 7 , crawley , perth , wa 6009 australia address = max planck institut fr radioastronomie , auf dem hgel 69 , 53121 bonn , germany address = csiro astronomy and space science , australia telescope national facility , csiro , po box 76 , epping , nsw 1710 , australia address = icrar / curtin institute of radio astronomy , gpo box u1987 , perth , wa 6845 australia address = csiro astronomy and space science , australia telescope national facility , csiro , po box 76 , epping , nsw 1710 , australia address = department of physics , mcgill university , 3600 university street , montreal , quebec , h3a 2t8 , canada address = netherlands institute for radio astronomy ( astron ) , postbus 2 , 7990 aa dwingeloo , the netherlands address = max planck institut fr radioastronomie , auf dem hgel 69 , 53121 bonn , germany , altaddress = university of manchester , jodrell bank centre for astrophysics , alan turing building , manchester m13 9pl , u.k .
address = jet propulsion laboratory , m / s 138 - 308 , 4800 oak grove dr . , pasadena , ca 91109 , usa address = jet propulsion laboratory , m / s 138 - 308 , 4800 oak grove dr . , pasadena ,
ca 91109 , usa address = csiro astronomy and space science , australia telescope national facility , csiro , po box 76 , epping , nsw 1710 , australia address = department of physics , lafayette college , easton , pa 18042 , usa address = inaf - osservatorio astronomico di cagliari , localit poggio dei pini , strada 54 , i-09012 capoterra , italy address = inaf - osservatorio astronomico di cagliari , localit poggio dei pini , strada 54 , i-09012 capoterra , italy address = national radio astronomy observatory , charlottesville , va 22903 , usa address = institut de cincies de lespai ( ieec - csic ) , campus uab , facultat de cincies , torre c5-parell , 2a planta , 08193 barcelona , spain address = csiro astronomy and space science , australia telescope national facility , csiro , po box 76 , epping , nsw 1710 , australia address = netherlands institute for radio astronomy ( astron ) , postbus 2 , 7990 aa dwingeloo , the netherlands , altaddress = university of manchester , jodrell bank centre for astrophysics , alan turing building , manchester m13 9pl , u.k .
address = university of manchester , jodrell bank centre for astrophysics , alan turing building , manchester m13 9pl , u.k .
address = icrea & institut de cincies de lespai ( ieec - csic ) , campus uab , facultat de cincies , torre c5-parell , 2a planta , 08193 barcelona , spain address = netherlands institute for radio astronomy ( astron ) , postbus 2 , 7990 aa dwingeloo , the netherlands , altaddress = astronomical institute `` anton pannekoek , '' university of amsterdam , 1098 sj amsterdam , netherlands address = swinburne university of technology , centre for astrophysics and supercomputing , mail h39 , po box 218 , vic 3122 , australia address = university of manchester , jodrell bank centre for astrophysics , alan turing building , manchester m13 9pl , u.k . |
double - barred galaxies are rather common systems in the universe , representing @xmath21/3 of all barred galaxies @xcite .
these structures are mainly seen by surface brightness profile decomposition , which requires not only a bulge and a disk but also two additional components with non - axial symmetry .
the presence of inner bars does not seem to correlate with the morphological type of the host galaxy and there is also no preferred angle between the two structures .
this property suggests that both bars rotate independently , as seen in observations and consistent with numerical simulations ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
inner bars play an important role in the evolution of galaxies as they might transport gas to the central regions , where they may trigger the formation of stars that , in turn , lead to the appearance of new structures such as disk - like components @xcite .
theoretically , this phenomenon can not be achieved with a single bar , since the flow of material would stop before reaching the galactic center ( e.g. at the inner lindblad resonance ) .
the presence of two embedded bars makes it possible and may even help to feed the central agn @xcite .
the study of the orbital make - up of different morphological components in double - barred galaxies is one of the best ways to understand their internal structure and dynamics .
nevertheless , this is a rather complex issue to address as there are three fundamental frequencies for the regular orbits ( i.e. two for the bars and one for the free oscillations ; * ? ? ?
* ) instead of the two frequencies as in the case of a single bar .
as a result , they do not have closed and periodic orbits , so @xcite introduced the _ loop _
concept to represent the stable orbits in a double - barred potential .
a more popular approach is the use of numerical simulations to describe the structures seen in real observations , such as in the recent works by @xcite and @xcite . in their simulations , these authors generate long - lived double bars with and without a dissipative component , respectively .
however , most existing observations have relied on long - slit spectroscopy along only a few position angles , which makes the comparison with the models somewhat difficult .
better suited integral - field spectroscopy has generally not been used to observe these systems , with the notable exceptions of @xcite and . here
we present ` sauron ` integral - field spectroscopy results on the stellar kinematics of four double - barred early - type galaxies .
we will focus on the detailed analysis of the stellar velocity dispersion maps , that reveal a peculiar feature , not seen before , along the extremes of the inner bar .
we present the observational evidence and discuss possible interpretations using numerical simulations . a more complete account of the ionised gas , stellar population properties and supporting numerical simulations will be the subject of a forthcoming paper .
integral - field spectroscopic data of four double - barred early - type galaxies were taken with the ` sauron ` spectrograph , attached to the william herschel telescope at the observatorio del roque de los muchachos ( la palma , spain ) .
the galaxies were selected from the catalog of , so that not only the lengths of the inner bars were a good match to the ` sauron ` field - of - view ( fov ) , but also allowed for enough fov in order to sample the transition region between the two bars .
we deliberately selected early - type galaxies so as to avoid the appearance of complex structures due to the presence of dust , and to make it easier to perform stellar population analysis .
we used the lr mode of ` sauron ` providing a @xmath3 fov with spatial sampling of @xmath4 per lens .
the spectral range covers the domain between 4800 - 5380 .
this setup produces 1431 spectra per pointing over the ` sauron ` fov , with a sampling of 1.1 pixel@xmath1 and a spectral resolution of 3.74 (fwhm ) .
the data reduction was performed using the xsauron package @xcite , following the same procedures outlined in @xcite .
total integration times range between 3 to 4 hours , providing a signal - to - noise ratio of @xmath5 at the centers of the galaxies .
additionally , in order to ensure the measurement of reliable stellar kinematics , we spatially binned our final datacubes using the voronoi 2d binning algorithm of @xcite , creating compact bins with a minimum signal - to - noise ratio ( @xmath6 ) of @xmath7 per spectral resolution element .
most spectra in the central regions , however , have @xmath6 in excess of @xmath8 , and so remain un - binned .
the stellar kinematics is then extracted as described in @xcite , by fitting the absorption spectrum to a linear combination of the same set of templates . for this purpose
we used the penalized pixel - fitting ( ppxf ) method of @xcite and a well selected subsample of the stellar population models from @xcite , with evenly sampled ages and metallicities .
figure [ fig : data ] shows the results from the analysis of the datacubes .
the stellar velocity maps look very much like those of non double - barred galaxies and in three out of the four galaxies ( ngc2859 , ngc4725 and ngc5850 ) we find local velocity maxima and minima along the direction of the kinematic major axes , at a few arcseconds from the galaxy centers .
moreover , the @xmath9 maps , not included in this letter , show a clear anticorrelation with the stellar velocities at the locations of these features , supporting the idea that these faster rotating components are kinematically decoupled inner disks @xcite .
since the kinematic major axes are not aligned with the inner bars at these regions , these disks can not be related with the bars . however , the stellar velocity dispersion maps reveal some interesting features not seen in other morphological types of galaxies .
the values measured are at a maximum at the centers of the galaxies ( as seen in most e / s0 galaxies ) but , instead of smooth negative velocity dispersion gradients towards the outer parts , we find two symmetrical regions where the velocity dispersion values drop significantly compared to their surroundings .
these @xmath0-hollows are located exactly along the major axes of the inner bars and they extend out to the edges , as we checked with several tests .
the amplitude of the hollows varies between 10 and 40 km s@xmath1 .
interestingly , neither the h@xmath10 or @xmath11oiii@xmath12 $ ] emission - line maps ( not presented here ) show any distinct features at the same locations of the inner bars .
the stellar velocity dispersion map for ngc5850 was previously shown by , but no hollows are seen in there ; we believe this is because the smaller fov of the mpfs instrument used in those obervations.-2
since the @xmath0-hollows of the four galaxies are observed exactly at the ends of the inner bars , it seems that they are related to the inner bar itself and not to any other structural or kinematical component .
the aim of this section is to investigate the origin of the @xmath0-hollows , so in the following we discuss some possible explanations for these observations .
stellar inner disks could , in principle , explain a decrease in velocity dispersion in the central regions of galaxies , as happens with the @xmath0-drops seen in .
however , the observed alignment of the @xmath0-hollows with the major axis of the inner bar can not be accounted for with the presence of an inner disk whose major axis is usually well aligned with the main photometric major axis of the galaxy .
moreover , the decrease in @xmath0 in our galaxies does not occur at the same locations as the @xmath0-drops seen by ; in fact we find that the velocity dispersion reaches a maximum value at the center .
the possibility of a gaseous inner disk is discarded because the effect we see in our data is purely stellar .
a young stellar component that has acquired the kinematics of the cold gas it was formed from could also explain the decreases in the velocity dispersion values .
this young population might not necessarily be associated with a different structural component ( e.g. a nuclear star forming ring ) .
we tested this hypothesis by performing a preliminary stellar population analysis of the different structures to investigate the presence of young stars ( an extensive stellar population analysis will be published elsewhere ) .
however , we do not find any evidence of the presence of a particularly young stellar population at the @xmath0-hollows locations .
we now focus on the immediate surroundings of the inner bar . in the central parts of these galaxies we have the combination of a component with typically high velocity dispersion ( i.e. a bulge ) and the inner bar with its ordered motion and thus a low @xmath0 .
we propose that the presence of the @xmath0-hollows is due to the contrast between the velocity dispersion of these two components .
since the velocity dispersion profile and the luminosity profile of the bulge decreases outwards , we can only see this effect in the outer parts of the inner bar , where the bulge is not totally dominating the flux .
therefore the amplitude of the hollows will depend on the relative contribution of the bulge to the total luminosity at the extremes of the bar and on the difference of velocity dispersions between the bulge and the inner bar .
irrespective of the relative luminosity of the two components at the edges of the inner bar , if there were no differences between the velocity dispersions we would not be able to find any @xmath0-hollows .
assuming that there is a significant difference in the @xmath0 values at these points , a very extended bulge may have the inner bar embedded in it , so we would expect very deep @xmath0-hollows in this case . on the contrary , if the bulge were less extended these @xmath0-hollows would not show up . to test this idea we used the code ftm 4.4 ( updated version ) from @xcite to perform n - body self - consistent three - dimensional numerical simulations starting with a ( classical - like ) bulge and an exponential disk .
first , we relax the bulge , in order to have a `` quiescent '' start , and then we introduce an exponential self - graviting rotating disk . the bulge proceeds to evolve almost steadily , contrary to the disk , which develops a bar instability . here
, we use simulations with a single bar to mimic the main components in the region of interest : the central kiloparsec .
this simple approximation is enough to reproduce the @xmath0-hollows : they appear at the ends of the bar for the cases in which the contrast is sufficient .
these @xmath0-hollows last as long as the bar stays with a low @xmath0 .
stellar bars in general could be heated - up by the buckling instability , which is milder in the presence of gas @xcite and in weak bars @xcite .
since the four observed inner bars presented here are not very strong and there is some gas in these regions , most likely they will not buckle .
therefore , we expect the @xmath0-hollows to be long - lived structures . in figure
[ fig : model ] we illustrate two of our models : one case of a bulge large enough to match the length of the bar , where we are able to reproduce the @xmath0-hollows , and one case with a smaller , less extended bulge where the velocity dispersion profile is almost unchanged .
the simulations shown here use a galaxy inclination of 30@xmath13 ( as a representative case of our observations ) and a position angle of 20@xmath13 for the bar .
however , we repeated the tests with different observational conditions , getting similar results in all cases . note however that recent numerical simulations of double - barred systems by @xcite do not show any evidence of the presence of the @xmath0-hollows in their velocity dispersion maps .
a possible explanation for this absence is the fact that in those simulations there is not sufficient velocity dispersion difference between the resulting structural components .
we have presented 2d kinematical maps of double - barred early - type galaxies covering the whole nuclear region . we have found @xmath0-hollows appearing at the ends of the inner bar for our four objects .
the main result presented in this paper is that these hollows are signatures of inner bars , which indicates that it they must have significantly lower velocity dispersions than the surrounding bulges .
this means that inner bars are cold systems , and are not related to triaxial bulges .
moreover , the observations presented here put constraints on the degree of rotational over pressure support for these structural components .
finally , since the @xmath0-hollows are not due to other structural or kinematical components ( e.g. inner disks ) , they may be used to identify inner bars from a purely stellar kinematic analysis.-2 it is likely that previous numerical simulations were unable to predict the observed @xmath0-hollows because they did not include a sufficiently dynamically hot component ( i.e. classical bulge ) .
however , we have been able to reproduce this feature with simulations including a hot bulge , a single small bar and a disk .
the @xmath0-hollows can be seen due to a contrast between the high velocity dispersion of the bulge and the low velocity dispersion of the bar .
the amplitude of the hollows depends on the relative contributions to the total flux and the relative sizes of the bulge and the inner bar .
while young stellar populations can help to lower the stellar velocity dispersion , no clear signatures of them are found in our sample . based on the pieces of evidence presented in this paper
, we think that the @xmath0-hollows are simply a matter of contrast .
+ the authors thank e.emsellem , r.mcdermid and r.peletier for their help concerning data reduction and analysis .
their comprehensive comments , together with those from j.a.l.aguerri , p.t.dezeeuw , j.knapen and i.shlosman , have greatly helped to improve earlier drafts of the manuscript .
the authors are indebted to m.beasley for his help correcting the english , to the ` sauron ` team for their support for the realisation of this project and to the anonymous referee for constructive comments .
adlc gratefully acknowledges partial financial support from esa / estec and a european union eara fellowship for visiting the leiden observatory . this work has been supported by the grants aya2004 - 03059 and aya2007 - 67752-c03 - 01 , funded by the spanish ministry of education and science .
-hollows at the edges of the four inner bars ; these hollows are clearly seen in the velocity dispersion maps of ngc2859 , ngc4725 and ngc5850 whereas they are less evident ( that is , they have lower amplitudes and their presence is evident from cuts along the inner bar ) for the case of ngc3941.,width=453 ] -hollows . on the right
is the case of a very small bulge , whose luminosity drops quickly with radius such that its contribution is not relevant at the ends of the bar .
the @xmath0-hollows in this case become difficult to detect.,width=302 ] | we present ` sauron ` integral - field stellar velocity and velocity dispersion maps for four double - barred early - type galaxies : ngc2859 , ngc3941 , ngc4725 and ngc5850 .
the presence of the inner bar does not produce major changes in the line - of - sight velocity , but it appears to have an important effect in the stellar velocity dispersion maps : we find two _ @xmath0_-hollows of amplitudes between 10 and 40 km s@xmath1 on either side of the center , at the ends of the inner bars .
we have performed numerical simulations to explain these features .
ruling out other possibilities , we conclude that the _ @xmath0_-hollows are an effect of the contrast between two kinematically different components : the high velocity dispersion of the bulge and the more ordered motion ( low velocity dispersion ) of the inner bar . |
control problems in multi - agent systems have been attracting attention in diverse contexts @xcite@xcite . in the consensus problem , for example , the objective is to make all agents converge to some common state by designing proper algorithms @xcite-@xcite , such as the linear consensus protocol @xmath0 here , @xmath1 is the state of agent @xmath2 and @xmath3 are the components of the laplacian matrix @xmath4 , satisfying @xmath5 for all @xmath6 and @xmath7 .
the laplacian is associated with the underlying graph @xmath8 , whose links can be directed and weighted . it can be shown that , if the underlying graph has a spanning tree , then all agents converge to a common number , which depends on the initial values @xcite . on the other hand ,
if it is desired to steer the system to a prescribed consensus value , auxiliary control strategies are necessary . among these , _ pinning control _
is particularly attractive because it is easily realizable by controlling only a few agents , driving them to the desired value @xmath9 through feedback action : @xmath10 where @xmath11 denotes the subset of agents where feedback is applied , with cardinality @xmath12 , @xmath13 is the indicator function ( 1 if @xmath14 and 0 otherwise ) , and @xmath15 is the pinning strength . eq .
( [ pin ] ) provides the local strategy that pins a few nodes to stabilize the whole network at a common desired value .
the following hypothesis is natural in pinning problems and assumed in this paper . *
( h ) * each strongly connected component of @xmath16 without incoming links from the outside has at least one node in @xmath11 .
the following result is proved in @xcite .
[ thm0 ] if ( h ) holds , then system ( [ pin ] ) is asymptotically stable at @xmath17 @xmath18 . in many networked systems ,
however , time delays inevitably occur due to limited information transmission speed ; so proposition [ thm0 ] does not apply . in this paper
we consider systems with both transmission and pinning delays , @xmath19 for @xmath20 , where @xmath21 denotes the _ transmission delay _ in the network and @xmath22 is the _ pinning delay _ of the controllers .
several recent papers have addressed the stability of consensus systems with various delays .
it has been shown that consensus can be achieved under transmission delays if the graph has a spanning tree @xcite-@xcite .
however , if a sufficiently large delay is present also in the self - feedback of the node s own state , then consensus may be destroyed @xcite ; similar conclusions also hold in cases of time - varying topologies @xcite@xcite and heterogeneous delays @xcite-@xcite .
the stability of pinning networks with nonlinear node dynamics have been studied in @xcite@xcite , @xcite@xcite .
however , the role of pinning delay was considered in only a few papers @xcite@xcite , where it was argued that stability can be guaranteed if the pinning delays are sufficiently small .
precise conditions on the pinning delay for stability , the relation to the network topology , and the selection of pinned nodes have not yet been addressed . in this paper , we study the stability of the model ( [ pintransdelay ] ) under both transmission and pinning delays .
first , we derive an estimate of the largest admissible pinning delay .
next , we consider several specific scenarios and present numerical algorithms to verify stability by calculating the dominant eigenvalue of the system
. included among the scenarios are the cases when only a single node is pinned in the absence of transmission delay , or when the transmission and pinning delays are identical .
finally , we use a perturbation approach to estimate the dominant eigenvalue for very small and very large pinning strengths .
a directed graph @xmath23 consists of a node set @xmath24 and a link set @xmath25 .
a ( directed ) _ path _ of length @xmath26 from node @xmath27 to @xmath28 , denoted @xmath29 , is a sequence of @xmath30 distinct vertices with @xmath31 and @xmath32 such that @xmath33 for @xmath34 .
the graph is called strongly connected if there is a directed path from any node to any other node , and it is said to have a spanning tree if there is a node @xmath35 such that for any other node @xmath36 there is a path from @xmath37 to @xmath38 .
we denote the imaginary unit by @xmath39 and the @xmath40 identity matrix by @xmath41 . for a matrix @xmath4
, @xmath3 denotes its @xmath42 element and @xmath43 its transpose .
the laplacian matrix @xmath4 is associated with the graph @xmath16 in the sense that there is a link from @xmath27 to @xmath28 in @xmath8 if and only if @xmath44 .
we denote the eigenvalues of @xmath4 by @xmath45 .
recall that zero is always an eigenvalue , with the corresponding eigenvector @xmath46^\top$ ] , and @xmath47 for all nonzero eigenvalues @xmath48 . furthermore ,
if the graph @xmath8 is strongly connected ( or equivalently , if @xmath4 is irreducible ) , then zero is a simple eigenvalue of @xmath4 .
the diagonal element @xmath49 is the _ weighted in - degree _ of node @xmath2 .
let @xmath50 be the diagonal matrix of in - degrees and @xmath51 .
let @xmath52 , @xmath53^{\top}$ ] , and @xmath54 with @xmath55 .
system ( [ pintransdelay ] ) can be rewritten as @xmath56 considering solutions in the form @xmath57 with @xmath58 and @xmath59 , the characteristic equation of ( [ pintransdelaym ] ) is obtained as @xmath60=0.\label{chpintransdelay}\]]the asymptotic stability of ( [ pintransdelaym ] ) is equivalent to all characteristic roots @xmath61 of ( [ chpintransdelay ] ) having negative real parts .
the root having the largest real part will be termed as the dominant root or the dominant eigenvalue . for the undelayed case ,
proposition [ thm0 ] can be equivalently stated as follows .
[ cor0 ] if ( h ) holds , then all eigenvalues of @xmath62 have negative real parts .
we also state an easy observation for later use : [ lem0 ] for any two column vectors @xmath63 , @xmath64
we first show that the system ( [ pintransdelaym ] ) is stable for all values of the pinning delay @xmath65 smaller than a certain value @xmath66 .
[ thm1 ] assume condition ( h ) .
let @xmath67\ ] ] and define @xmath68 if @xmath69 , then system ( [ pintransdelaym ] ) is stable for all @xmath70 .
first , we take @xmath71 and prove stability for all @xmath72 .
assume for contradiction that there exists some characteristic root @xmath73 of ( [ chpintransdelay ] ) such that @xmath74 .
applying the gershgorin disc theorem to , we have @xmath75 for some @xmath2 , which implies @xmath76^{2}+[{{\mathrm{im}}}(\lambda^{*})]^{2}\le l_{ii}^{2}. \label{disc1}\ ] ] since @xmath77 , it must be the case that @xmath78 ; i.e. , @xmath79
. then @xmath80 , and since @xmath81 , gives @xmath82 .
this , however , contradicts corollary [ cor0 ] .
therefore , when @xmath71 , all characteristic roots of ( [ chpintransdelay ] ) have negative real parts .
we now let @xmath83 .
suppose ( [ chpintransdelay ] ) has a purely imaginary root @xmath84 , @xmath85 . by ( [ disc0 ] )
, we have , for some index @xmath86 , @xmath87 implying @xmath88^{2}+[\omega - cd_{q } \sin(\omega\tau_{p})]^{2}}\le l_{qq}.\ ] ] thus , @xmath89 we claim that @xmath86 must be a pinned node . for
if @xmath90 , then @xmath91 must be zero , which implies that zero is a characteristic root of ( [ chpintransdelay ] ) , contradicting corollary [ cor0 ] .
therefore @xmath92 . in the notation of ,
the inequality can then be written as @xmath93 . by ( [ taup ] ) , however , we have that @xmath94 for all @xmath95 , @xmath96 and @xmath69 . we conclude that ( [ chpintransdelay ] ) does not have purely imaginary roots for @xmath69 .
thus , by ( * ? ? ?
* theorem 2.1 ) , all characteristic roots of ( [ chpintransdelay ] ) have strictly negative real parts for @xmath69 .
proposition [ thm1 ] provides an estimate for the largest admissible pinning delay for which system ( [ pintransdelaym ] ) is stable .
this estimate needs only the knowledge of the set of pinned nodes and their weighted in - degrees .
we now consider the possibility of controlling the network using a single node , say , the @xmath86th one .
then @xmath97 , where @xmath98 denotes the @xmath86th standard basis vector , whose @xmath86th component is one and other components zero .
if @xmath99 is nonsingular , the characteristic equation ( [ chpintransdelay ] ) becomes @xmath100 \nonumber\\ = & \det(\lambda i_{n}+k - a\exp(-\lambda\tau_{r}))\nonumber\\ & \det\left[i_{n}+cu_{q}u_{q}^{\top}(\lambda i_{n}+k - a\exp(-\lambda\tau_{r}))^{-1}\exp(-\lambda\tau_{p})\right]\nonumber\\ = & \det(\lambda i_{n}+k - a\exp(-\lambda\tau_{r}))\nonumber\\ & ( 1+c u_{q}^{\top}(\lambda i_{n}+k - a\exp(-\lambda\tau_{r}))^{-1}u_{q}\exp(-\lambda\tau_{p}))\label{eqnxx1}\end{aligned}\ ] ] using lemma [ lem0 ] .
then we have the following result . [ prop3 ]
assume ( h ) .
if all solutions @xmath61 of the equation @xmath101 satisfy @xmath102 , then system ( [ pintransdelaym ] ) is stable . as in the first part of the proof of proposition [ thm1 ] , the equation @xmath103=0 $ ] has no solutions with @xmath104 .
hence , if all solutions @xmath61 of ( [ eqnx1 ] ) have negative real parts , then all roots of ( [ chpintransdelay ] ) have negative real parts .
we consider two specific cases to obtain more information about the solutions of .
first , we consider the absence of transmission delays , i.e. , @xmath105 .
suppose for simplicity that @xmath4 is diagonalizable and has only real eigenvalues : @xmath106 for some nonsingular @xmath107 and a real diagonal matrix @xmath108 of eigenvalues of @xmath4 .
the column vectors of @xmath109 ( resp , the row vectors of @xmath107 ) are the right ( resp . , left )
eigenvectors of @xmath4 .
then , ( [ eqnx1 ] ) can be written as @xmath110 where @xmath111 is the @xmath86th left eigenvector and @xmath112 is the @xmath86th right eigenvector of @xmath4 .
we expand as @xmath113 in terms of the components @xmath114 of @xmath115 and @xmath116 , respectively .
consider the smallest value of @xmath22 for which there exists a purely imaginary solution , @xmath84 .
then , the real and imaginary parts of ( [ chsing ] ) give @xmath117 where @xmath118 rearranging gives @xmath119 and @xmath120 .
this implies @xmath121 and @xmath122 we then have the following result . [ single_pin_thm1 ] suppose @xmath105 , @xmath4 is diagonalizable , irreducible , and all its eigenvalues are real .
let the eigenvalues @xmath123 of @xmath4 be sorted so that @xmath124 , and let @xmath125 $ ] , @xmath126 , be the left eigenvector of @xmath4 corresponding to the zero eigenvalue .
let @xmath127 denote the set of positive solutions of the equation @xmath128 with respect to the variable @xmath129 , where @xmath130 and @xmath131 are given by ( [ ab ] ) .
define @xmath132 then system ( [ pintransdelaym ] ) is stable for @xmath133 .
( [ eqnxx1 ] ) implies that any purely imaginary solution @xmath134 of ( [ chpintransdelay ] ) should also be a solution of ( [ chsing ] )
. then @xmath91 must be a real solution of ( [ eqnx2 ] ) . by the definition of @xmath127 , the solution set of ( [ eqnx2 ] ) with respect to @xmath91 is @xmath135 . by the assumption of irreducibility , @xmath136 for all @xmath137 and @xmath138 @xmath18 . if @xmath139 , then the smallest positive solution of ( [ tt ] ) with respect to @xmath22 is @xmath140 . if , on the other hand , @xmath141 , noting that @xmath142 and @xmath143 , the smallest positive solution of ( [ tt ] ) is again @xmath140 . therefore , given @xmath144 , the smallest nonnegative solution of ( [ tt ] ) with respect to @xmath22 should be in the set @xmath145 . since the mapping @xmath146 is a decreasing function of @xmath147 , the quantity @xmath148 defined in ( [ taupp ] ) is the smallest nonnegative solution of ( [ tt ] ) with respect to @xmath22 , given @xmath144
hence , for @xmath149 ( [ chsing ] ) does not have any purely imaginary solutions . since for @xmath71 all characteristic roots of ( [ chpintransdelay ] ) have negative real parts , we conclude that all roots have negative real parts for @xmath149 . by derivation ,
( [ chsing ] ) is independent of the ordering of the eigenvalues or the eigenvectors in @xmath150 .
therefore , the bound @xmath148 for allowable pinning delays given in proposition [ single_pin_thm1 ] does not depend on the choice of the pinned node .
proposition [ single_pin_thm1 ] suggests an algorithm to calculate @xmath148 : 1 .
find the largest positive solution @xmath151 of the equation @xmath152 2 .
calculate ( [ taupp ] ) .
we illustrate this approach in an erds - renyi ( e - r ) random network of @xmath153 nodes with linking probability @xmath154 , where the first node is pinned .
the left and right eigenvectors of @xmath4 associated with the zero eigenvalue are given by @xmath155/\sqrt{n}$ ] .
figure [ single_region ] shows the parameter region @xmath156 , illustrating the inverse dependence of @xmath157 on @xmath158 .
note that @xmath159 does not necessarily imply instability , since proposition [ single_pin_thm1 ] gives only a sufficient condition .
nevertheless , the curve shown in fig . [ single_region](a ) turns out to be a good approximation of the boundary of the exact stability region .
to illustrate , we take two parameter points very close ( @xmath160 of the @xmath148 ) to the curve but on different sides of it , as indicated by blue and red stars in fig . [
. we simulate ( [ pintransdelay ] ) at the corresponding parameter values , with the same laplacian as above and @xmath105 . as seen in fig .
[ single_region](b)(c ) , the two points indeed yield different stability properties .
the other situation we consider is the homogeneous case when @xmath4 is diagonalisable and normalised , i.e. , @xmath161 @xmath18 for some @xmath162 , and @xmath163 .
then ( [ eqnx1 ] ) becomes @xmath164 let @xmath165 ; thus @xmath166 .
then , by the same algebra as above , ( [ eqnx12 ] ) becomes @xmath167 we have the following result . [ single_pin_thm2 ] suppose that @xmath163 , @xmath4 is diagonalizable , irreducible , normalised ( @xmath161 @xmath18 ) , and all its eigenvalues @xmath168 are real .
denote @xmath124 and let @xmath125 $ ] be the left eigenvector of @xmath4 corresponding to the eigenvalue @xmath169 , with @xmath170 .
let @xmath171 denote the set of all the branches of the solutions of the equation @xmath172 with respect to the variable @xmath9 .
then system ( [ pintransdelaym ] ) is stable whenever the real parts of the numbers @xmath173 are all negative , where @xmath174 is the lambert @xmath174 function @xcite .
proposition [ single_pin_thm2 ] can be proved by transforming ( [ chsing1 ] ) into ( [ eqnx22 ] ) with @xmath175 and using proposition [ prop3 ] .
in this section , we consider the extreme situations when the pinning strength @xmath158 is very small or very large .
we will employ the perturbation approach in @xcite to approximate the eigenvalues and eigenvectors in terms of @xmath158 .
the characteristic roots @xmath61 of ( [ chpintransdelay ] ) are eigenvalues of the matrix @xmath176 .
hence , when @xmath177 , the characteristic roots of ( [ chpintransdelay ] ) equal to the eigenvalues @xmath178 of @xmath179 . under the condition ( h )
, there is a single eigenvalue @xmath180 .
we denote the right and left eigenvectors of @xmath181 by @xmath182 and @xmath183 respectively , with @xmath184 .
it can be seen that @xmath185 and @xmath186 ( associated with @xmath180 ) are , respectively , the right and left eigenvectors of @xmath4 associated with the zero laplacian eigenvalue .
let @xmath187 denote the characteristic roots of ( [ chpintransdelay ] ) and @xmath188 and @xmath189 denote the right and left eigenvectors of @xmath190 , regarded as functions of @xmath158 , with @xmath191 , @xmath192 and @xmath193 . using a perturbation expansion @xcite , @xmath194 where @xmath195 denotes terms that satisfy @xmath196 .
thus , @xmath197\tilde{\phi}^{i}(c)\\ & & = \lambda_{i}(c)\tilde{\phi}^{i}(c).\end{aligned}\ ] ] when @xmath158 is sufficiently small , the dominant eigenvalue is @xmath198 , since @xmath180 is the dominant eigenvalue when @xmath177 .
hence , we consider @xmath199
. then @xmath200 . comparing the first - order terms in @xmath158 on both sides , @xmath201 . multiplying both sides with @xmath202 and noting that @xmath203 , @xmath204 hence , we have the following result . [ thm4 ]
suppose that the underlying graph is strongly connected and at least one node is pinned .
then , for sufficiently small @xmath158 , all characteristic roots of ( [ chpintransdelay ] ) have negative real parts and the dominant root is given by @xmath205 since the graph is strongly connected , @xmath4 has a simple zero eigenvalue . when @xmath177 , the dominant root of ( [ chpintransdelay ] ) is @xmath206 .
since the roots of ( [ chpintransdelay ] ) depend analytically on @xmath158 , they are given by @xmath198 for all sufficiently small @xmath158 .
substituting ( [ lambda11 ] ) into @xmath198 and noting that @xmath207 completes the proof . in order to understand the meaning of ( [ approxc ] ) , consider the special case of an undirected graph with binary adjacency matrix @xmath208 .
then , with @xmath209^{\top}$ ] and @xmath210^{\top}/n$ ] , we have @xmath211 , which equals the _ average degree _ of the graph . in addition
, @xmath212 , which is the _ fraction of pinned agents_. then , ( [ approxc ] ) yields the approximation @xmath213 for small @xmath158 , which uses only the pinning fraction and the mean degree of the graph .
since the real part of the dominant characteristic value measures the exponential convergence of the system , proposition [ thm4 ] implies that , for sufficiently small @xmath158 , the convergence rate is improved if the number of pinned nodes is increased , the transmission delay is reduced , or the mean degree is decreased .
if the graph is directed , a similar statement can be obtained by taking the components of @xmath185 as weights : @xmath214 . to illustrate this result
, we employ a numerical method to calculate the real part of @xmath198 , namely , by simulating the system and expressing its exponential convergence rate in terms of its largest lyapunov exponent . in detail , letting @xmath215 , we partition time into disjoint intervals of length @xmath216 , @xmath217 , and define @xmath218 for @xmath219 $ ] .
then , the largest lyapunov exponent , which equals to the largest real part of solutions of ( [ chpintransdelay ] ) , is numerically calculated via @xcite @xmath220 where @xmath221 stands for the function norm .
the latter is numerically calculated by approximating @xmath222 with a finite - dimensional vector @xmath223 obtained by evaluating @xmath224 at a finite number of equally spaced points and using the vector norm @xmath225 .
the estimate can then be compared with the analytical estimate for @xmath226 obtained from ( [ approxc ] ) : @xmath227 for simulations , we generate an undirected e - r random graph of @xmath153 nodes with linking probability @xmath228 and randomly select a given fraction @xmath229 of them as the pinned nodes .
the pinning delay is taken as @xmath230 .
figure [ fig4_small_c ] shows that the simulated value of @xmath226 decreases almost linearly with respect to @xmath158 and @xmath229 , and increases with respect to @xmath21 and the mean degree .
the simulation results are in a good agreement with the theoretical results .
the error between @xmath231 and @xmath232 depends on the values of @xmath233 and @xmath158 .
it can be seen that the error will increase as @xmath158 or @xmath233 ( or equivalently , @xmath229 ) increases , or else as the mean degree or @xmath21 decreases .
next , we consider the case of large @xmath158 . letting @xmath234 and @xmath235 , ( [ chpintransdelay ] ) is rewritten as @xmath236=0 .
\label{large_tau_ch}\ ] ] by the foregoing results , one can see that when @xmath237 is sufficiently small , equivalently , @xmath158 is sufficiently large , the largest admissible pinning delay for approaches zero .
it is therefore natural to assume that @xmath22 depends on @xmath158 in such a way that @xmath238 is bounded as @xmath158 grows large .
thus , we assume that @xmath239 remains bounded as @xmath240 .
when @xmath241 , ( [ large_tau_ch ] ) becomes approximately @xmath242 , where @xmath243 can be any value between @xmath244 and @xmath245 . in terms of components , @xmath246 if @xmath247 , and @xmath169 otherwise . the characteristic equation ( [ large_tau_ch ] ) with @xmath241 can be written as @xmath248 where @xmath249 .
it is known that @xmath250 for all roots of the function @xmath251 if and only if @xmath252 .
therefore , we impose the condition : @xmath253 . thus , the largest real part of the solutions of ( [ large_tau_ch_esp_0 ] ) is zero , and is obtained for the solution @xmath254 .
the corresponding eigenspace has dimension @xmath255 and has the form @xmath256^{\top}\in\mathbb
r^{n}:~u_{i}=0,~\forall~i\in\mathcal d\}.\ ] ] without loss of generality , we assume @xmath257 .
thus , we consider perturbation in terms of @xmath237 near zero eigenvalues @xmath258 and its corresponding right and left vectors , @xmath259 such that @xmath260 and @xmath261 if @xmath6 , @xmath262 .
let @xmath263 stand for the perturbed solution of ( [ large_tau_ch ] ) , @xmath264 and @xmath265 be the corresponding right and left eigenvectors , respectively . by a perturbation expansion , @xmath266 as @xmath267 .
thus , from ( [ large_tau_ch ] ) , @xmath268\tilde{\xi}^{i}(\epsilon)\\ & & = \mu_{i}(\epsilon)\tilde{\xi}^{i}(\epsilon).\end{aligned}\ ] ] since @xmath269 , by comparing the coefficients of order @xmath270 , we have @xmath271\xi^{i}-d\xi^{i,1}=\mu^{1}_{i}\xi^{i}.\label{large_order_1}\end{aligned}\ ] ] we write @xmath272 , \ ;
a=\left[\begin{array}{ll}a_{11}&a_{12}\\ a_{21}&a_{22}\end{array}\right ] , \ ; d=\left[\begin{array}{ll}i_{m}&0\\ 0&0\end{array}\right],\ ] ] and @xmath273 , @xmath274^{\top}$ ] , @xmath275^{\top}$ ] , with @xmath276 , @xmath277 , @xmath278 and @xmath279 corresponding to the pinned subset @xmath11 of dimension @xmath280 .
then ( [ large_order_1 ] ) becomes @xmath281\xi^{i}_{2 } & = \mu^{1}_{i}\xi^{i}_{2}\\ \exp(-\mu^{1}_{i}\tau_{r } ) a_{12}\xi^{i}_{2}-\xi^{i,1}_{1 } & = 0 .
\end{cases } \label{large_order_1x}\ ] ] we have the following result . [ thm5 ] suppose that the underlying graph is strongly connected and at least one node is pinned .
fix @xmath70 , and suppose @xmath282 as @xmath283
. then the dominant root of ( [ large_tau_ch ] ) has the form @xmath284 where @xmath285 is the dominant eigenvalue of the delay - differential equation @xmath286 furthermore , @xmath287 for all sufficiently large @xmath158 .
the condition @xmath288 implies that , when @xmath241 , the dominant root of the characteristic equation ( [ large_tau_ch ] ) is zero and corresponds to the eigenspace @xmath289 .
so , for sufficiently small @xmath237 , the dominant root of equation ( [ large_tau_ch ] ) and the corresponding eigenvector have the form ( [ large_perturb ] ) , where @xmath290 satisfies the first equation in ( [ large_order_1x ] ) , i.e. , is an eigenvalue of ( [ large_c_equ ] ) . since @xmath291 , ( [ lambda_large_c ] ) follows .
moreover , since @xmath292 is diagonally dominant , one can see that @xmath293 under condition ( h ) .
therefore , for sufficiently large @xmath158 , all characteristic values of system ( [ pintransdelay ] ) have negative real parts .
we note that @xmath294 depends only on the coupling structure of the uncoupled nodes . to illustrate this result , we consider examples with a similar setup as in sec . [ small_pinning ] .
we take an e - r graph with @xmath153 nodes and linking probability @xmath228 , and pin @xmath295 nodes .
we set @xmath296 and @xmath297 .
the real part of the dominant characteristic root of ( [ chpintransdelay ] ) is numerically calculated via the largest lyapunov exponent , using formula ( [ le ] ) .
its theoretical estimation comes from theorem [ thm5 ] : @xmath298 , where the largest real part of @xmath299 is similarly calculated from the largest lyapunov exponent of ( [ large_c_equ ] ) .
[ fig5 ] shows that as @xmath158 grows large , the real part of the dominant root of ( [ chpintransdelay ] ) obtained from simulations approach the theoretical result @xmath300 , thus verifying proposition [ thm5 ] .
we have shown in this paper that the stability of the multi - agent systems with a local pinning strategy and transmission delay may be destroyed by sufficiently large pinning delays .
using theoretical and numerical methods , we have obtained an upper - bound for the delay value such that the system is stable for any pinning delay less than this bound . in this case , the exponential convergence rate of the multi - agent , which equals the smallest nonzero real part of the eigenvalues of the characteristic equation , measures the control performance .
99 m. h. degroot , reaching a consensus , j. amer .
assoc . , 69 ( 1974 ) , 118121 . c. reynolds , flocks , herds , and schools : a distributed behavioral model , comput .
, 21:4 ( 1987 ) , 2534 .
t. vicsek , a. czirk , e. ben - jacob , i. cohen and o. shochet , novel type of phase transition in a system of self - driven particles , phys .
lett . , 75 ( 1995 ) , 12261229 . t. p. chen , x. w. liu , w. l. lu .
pinning complex networks by a single controller .
ieee trans .
circuits syst .
i , 54:6 ( 2007 ) , 13171326 .
w. l. lu , x. li , z. h. rong .
global stabilization of complex networks with digraph topologies via a local pinning algorithm .
automatica , 46 ( 2010 ) , 116121 . q. song , f. liu , j. cao , w. yu , m - matrix strategies for pinning - controlled leader - following consensus in multiagent systems with nonlinear dynamics , ieee trans .
cybern . , 43:6 ( 2013 ) , 16881697 .
r. olfati - saber and r. m. murray , consensus problems in networks of agents with switching topology and time - delays , ieee trans .
control , 49 ( 2004 ) , 15201533 .
f. m. atay , consensus in networks under transmission delays and the normalized laplacian .
a , 371 ( 2013 ) , 20120460 . f. m. atay , on the duality between consensus problems and markov processes , with application to delay systems .
markov processes and related fields ( in press ) . p .- a .
bliman and g. ferrari - trecate , average consensus problems in networks of agents with delayed communications , automatica , 44 ( 2008 ) , 19851995 .
l. moreau , stability of multiagent systems with time - dependent communication links , ieee trans .
control , 50 ( 2005 ) , 169182 .
f. xiao and l. wang , asynchronous consensus in continuous - time multi - agent systems with switching topology and time - varying delays , ieee trans .
control , 53 ( 2008 ) , 18041816 .
w. l. lu , f. m. atay , j. jost , consensus and synchronization in discrete - time networks of multi - agents with stochastically switching topologies and time delays , networks and heterogeneous media , 6:2(2011 ) , 329349 .
m. ulrich , a. papachristodoulou , f. allgwer , generalized nyquist consensus condition for linear multi - agent systems with heterogeneous delays , proc .
ifac workshop on estimation and control of networked systems , 2429 , 2009 .
s. ruan , j. wei . on the zeros of transcendental functions with applications to stability of delay differential equations with two delays .
. discrete impuls .
systems , ser .
a 10 ( 2003 ) , 863874 . | we study the stability of networks of multi - agent systems with local pinning strategies and two types of time delays , namely the transmission delay in the network and the pinning delay of the controllers .
sufficient conditions for stability are derived under specific scenarios by computing or estimating the dominant eigenvalue of the characteristic equation .
in addition , controlling the network by pinning a single node is studied .
moreover , perturbation methods are employed to derive conditions in the limit of small and large pinning strengths .
numerical algorithms are proposed to verify stability , and simulation examples are presented to confirm the efficiency of analytic results . |
in the historical conference _ combinatoire numerative _ that took place at the end of may 1985 , in montreal , richard stanley raised some intriguing problems about the enumeration of plane partitions ( see below ) , which he later expanded into a fascinating article @xcite .
most of these problems concerned the enumeration of `` symmetry classes '' of plane partitions that were discussed in more detail in another article of stanley @xcite .
all of the conjectures in the latter article have since been proved ( see david bressoud s modern classic @xcite ) , except one , which until now resisted the efforts of the greatest minds in enumerative combinatorics .
it concerns the proof of an explicit formula for the @xmath0-enumeration of totally symmetric plane partitions , conjectured , ca .
1983 , independently by george andrews and david robbins ( @xcite , @xcite conj . 7 , @xcite conj
. 13 , and already alluded to in @xcite ) . in the present article
we finally turn this conjecture into a theorem .
a plane partition @xmath1 is an array @xmath2 , of positive integers @xmath3 with finite sum @xmath4 , which is weakly decreasing in rows and columns so that @xmath5 and @xmath6 . a plane partition @xmath1 is identified with its 3d ferrers diagram which is obtained by stacking @xmath3 unit cubes on top of the location @xmath7 .
this gives a left- , back- , and bottom - justified structure in which we can refer to the locations @xmath8 of the individual unit cubes .
if the diagram is invariant under the action of the symmetric group @xmath9 then @xmath1 is called a totally symmetric plane partition ( tspp ) . in other words , @xmath1 is called totally symmetric if whenever a location @xmath8 in the diagram is occupied then all its up to @xmath10 permutations @xmath11 are occupied as well .
such a set of cubes , i.e. , all cubes to which a certain cube can be moved via @xmath9 is called an orbit ; the set of all orbits of @xmath1 forms a partition of its diagram ( see figure [ tsppfigure ] ) .
[ tsppfigure ] .
the yellow cubes form an orbit of length @xmath12 ( corresponding to all permutations of @xmath13 ) , the red cubes an orbit of length @xmath14 ( all permutations of @xmath15 ) , and the green cube on location @xmath16 represents an orbit of length @xmath17.,title="fig:",width=283 ] in 1995 , john stembridge @xcite proved ian macdonald s conjecture that the number of totally symmetric plane partitions with largest part at most @xmath18 , i.e. , those whose 3d ferrers diagram is contained in the cube @xmath19 ^ 3 $ ] , is given by the elegant product - formula @xmath20 ten years after stembridge s completely human - generated proof , andrews , peter paule and carsten schneider @xcite came up with a computer - assisted proof based on an ingenious matrix decomposition , but since no @xmath0-analog of their decomposition was found , their proof could not be extended to a proof for the @xmath0-case . a third proof of stembridge s theorem @xcite , completely computer - generated , was recently found in the context of our investigations of the @xmath0-tspp conjecture .
we have now succeeded in completing all the required computations for an analogous proof of the @xmath0-tspp conjecture , and can therefore announce : [ qtsppthm ] let @xmath21 denote the set of orbits of a plane partition @xmath1 under the action of the symmetric group @xmath9 . then the orbit - counting generating function ( @xcite , @xcite ) is given by @xmath22 where @xmath23 denotes the set of totally symmetric plane partitions with largest part at most @xmath18 .
our proof is based on a result by soichi okada @xcite who has shown that the theorem is implied by a certain conjectured determinant evaluation .
these preliminaries are stated accurately in section [ secreduction ] below , followed by a description of the holonomic ansatz @xcite that we are going to pursue .
this approach relies on a kind of oracle that tells us a description of a certificate function @xmath24 ; the odyssey how this function has been `` guessed '' is described in section [ secguess ] .
once @xmath24 is known , the determinant evaluation reduces further to proving the three identities , , and stated below . in section [ secproof ]
we come full circle by proving these identities .
additional technical details of the proof and our computations as well as the explicit certificates are provided electronically on our website ( http://www.risc.uni-linz.ac.at/people/ckoutsch/qtspp/ ) .
our proof is noteworthy not only for its obvious significance in enumerative combinatorics , as it settles a long - standing conjecture , attempted by many people .
it is noteworthy also for computational reasons , as the computations we performed went far beyond what has been thought to be possible with currently known algebraic algorithms , software packages , and computer hardware .
in order to prove the @xmath0-tspp conjecture , we exploit an elegant reduction by okada @xcite to the problem of evaluating a certain determinant . this determinant is also listed as conjecture 46 in christian krattenthaler s essay @xcite on the art of determinant evaluation .
let , as usual , @xmath25 be the kronecker delta function and let , also as usual , @xmath26 denote the @xmath0-binomial coefficient .
define the discrete function @xmath27 by : @xmath28 okada s crucial insight is that theorem [ qtsppthm ] holds if @xmath29 so for proving the @xmath0-tspp conjecture , it is sufficient to prove this conjectured determinant evaluation . for this purpose
we apply a computational approach originally proposed in @xcite which is applicable to identities of the form @xmath30 where @xmath27 and @xmath31 ( with @xmath32 for all @xmath33 ) are given explicitly ( as it is the case here ) .
the approach rests on the following induction argument on @xmath18 . for @xmath34 ,
the identity is trivial .
suppose the identity holds for @xmath35 ; then the linear system @xmath36 has a unique solution @xmath37 .
the component @xmath24 of this solution is precisely the @xmath38-cofactor divided by the @xmath39-cofactor of the @xmath40-determinant .
the division is meaningful because the @xmath39-cofactor is just the @xmath41-determinant , which by induction hypothesis is equal to @xmath42 , which by general assumption is nonzero . since the @xmath40-determinant can be expressed in terms of the matrix entries @xmath43 and the normalized cofactors @xmath24 via @xmath44 the induction step is completed by showing that this sum evaluates to @xmath31 .
the difficulty is that this last summation involves the function @xmath24 ( of the discrete variables @xmath18 and @xmath45 ) for which we do not have an explicit expression for general @xmath18 and @xmath45 . in order to achieve our goal
, we guess a suitable description of a function @xmath24 and then prove that it satisfies the three identities @xmath46 once this is done , then by the argument given before the @xmath24 must be precisely the normalized @xmath38-cofactors of the @xmath40-determinant and the determinant evaluation follows as a consequence .
so in a sense , the function @xmath24 plays the rle of a certificate for the determinant identity .
in our the setting the certificate function @xmath24 will be described implicitly by a system of linear recurrence equations in @xmath18 and @xmath45 with coefficients depending polynomially on @xmath0 , @xmath47 and @xmath48 .
such recurrence equations can be phrased as the elements of some noncommutative operator algebras @xmath49 $ ] where the symbols @xmath50 represent the shifts @xmath51 .
if a function is annihilated by certain operators ( viz.it satisfies certain recurrence equations ) , then it is also annihilated by all the elements in the ( left ) ideal generated by those operators .
we speak of an annihilating ideal and represent such ideals by ( left ) grbner bases , so that for instance ideal membership can be decided effectively .
the annihilating ideals we use for representing functions are such that they uniquely determine the function up to some finitely many initial values .
technically , this means that the ideals have dimension zero and that some particular polynomial coefficients appearing in the recurrence system must not vanish simultaneously for infinitely many points @xmath38 .
these recurrence systems are similar but somewhat simpler than @xmath0-holonomic systems @xcite , which satisfy some additional requirements that are not needed for our proof .
we have checked that all the ideals arising in our proof are indeed of the desired form , but in the interest of clarity we suppress a more detailed description of these checks here . for a complete analysis including all technical details we refer to the supplementary material on our website .
also in the interest of clarity , we will from now on identify recurrence equations with their corresponding operators .
a priori , there is no reason why the normalized cofactors @xmath24 should admit a recursive description of the kind we are aiming at , but there is also no reason why they should not .
it turns out that they do , and this is fortunate because for functions described in this way , techniques are known by which the required identities , and can be proven algorithmically @xcite . in order to find a recursive description for @xmath24
, we first computed explicitly the normalized cofactors @xmath52 for a few hundred specific indices
@xmath18 and @xmath45 by directly solving the linear system quoted in section [ secreduction ] . using an algorithm reminiscent of polynomial interpolation , we then constructed a set of recurrences compatible with the values of @xmath24 at the ( finitely many ) indices we computed .
polynomial interpolation applied to a finite sample @xmath53 of an infinite sequence @xmath54 will always deliver some polynomial @xmath55 of degree at most @xmath56 which matches the given data . if it turns out that this polynomial matches some further sequence terms @xmath57 , then it is tempting to conjecture that @xmath58 for all @xmath18 .
the more specific points @xmath18 are found to match , the higher is the evidence in favor of this conjecture , even though without further knowledge no finite amount of points will ever constitute a rigorous proof .
very much analogously , it is possible to extract recurrence equations from some finite number of values @xmath24 .
the equations become trustworthy if they also hold for points @xmath38 which were not used for their construction . in this way , we have discovered a system of potential recurrence equations for the @xmath24 , which , despite being respectable in size ( about 30 megabytes ) , appears to be a rather plausible candidate for a recursive description of the normalized cofactors @xmath24 .
the system is available for download on our website .
now we switch our point of view .
we discard the definition that @xmath24 be the normalized @xmath38-cofactor and redefine the discrete function @xmath24 as the unique solution of the guessed recurrence system whose ( finitely many ) initial values agree with the normalized @xmath38-cofactor .
if we succeed in proving that @xmath24 defined in this way satisfies the identities , and , then we are done . for this purpose
we provide operators which belong to the annihilating ideal of @xmath24 or related ideals and have certain features which imply the desired identities .
so in a sense , these operators play the rle of certificates for the identities under consideration .
because of their astronomical size ( up to 7 gigabytes ; equivalent to more than one million printed pages ; corresponding to about 2.5 tons of paper ) , these certificates are not included explicitly in this article but provided only electronically on our website . also because of their size , it was not possible to construct them by simply applying the standard algorithms from @xcite . a detailed explanation of how exactly we found the certificates is beyond the scope of this article and will be given in a separate publication .
but this lack of explanation does not at all affect the soundness of our proof , because the correctness of the certificates can be checked independently by simply performing ideal membership tests .
our certificates are so big that even this `` simple '' calculation is not quite trivial , but a reader with a sufficient amount of patience and programming expertise will be able to do it .
we proceed by explaining the properties of the certificates provided on our website and why they imply , and .
to certify that @xmath59 for all @xmath33 , we provide a recurrence in the annihilating ideal of @xmath24 which is of the special form @xmath60 by virtue of the substitution @xmath61 , it translates into a recurrence for the diagonal sequence @xmath62 .
this recurrence contains the annihilating operator @xmath63 of the constant sequence @xmath17 as a ( right ) factor , and therefore , after checking @xmath64 , it can be concluded that @xmath59 for all @xmath18 .
similarly , we showed that @xmath65 for all @xmath33 and that @xmath66 for all @xmath67 .
this knowledge greatly simplifies the following proofs of the summation identities . in order to prove the first summation identity , we translate into the equivalent formulation @xmath68 taking into account that @xmath65 .
we provide certificates for . to this end , let @xmath69 be the summand of the sum on the left - hand side .
a recursive description for @xmath69 can be computed directly from the defining equations of @xmath24 and the fact that the rest of the summand is a @xmath0-hypergeometric factor . in the corresponding operator
ideal @xmath70 we were able to find two different recurrence equations for @xmath69 which are of the special form @xmath71 where the @xmath72 are rational functions in @xmath73 and @xmath74 is a @xmath75-linear combination of certain shifts of @xmath69 ( which are determined by the grbner basis of @xmath70 as described in @xcite ) .
next observe that @xmath76 for @xmath77 ( because of the @xmath0-binomial coefficient ) and also for @xmath67 ( because @xmath66 for @xmath67 ) .
when summing the two recurrence equations for @xmath45 from @xmath78 to @xmath79 , the right - hand side telescopes to @xmath80 and the left - hand side turns into a recurrence for the sum in .
this method is known as creative telescoping .
finally we obtain two annihilating operators @xmath81 and @xmath82 for the left - hand side of . for the right - hand side of
, we can again construct an ideal of recurrences from the defining equations of @xmath24 .
it turns out that this ideal contains @xmath81 and @xmath82 , so that both sides of are annihilated by these two operators .
additionally , @xmath81 and @xmath82 have been constructed such that they require only finitely many initial values to produce a uniquely determined bivariate sequence .
therefore it suffices to check that the two sides of agree at some finitely many points @xmath38 . to certify the final identity , we rewrite equivalently into @xmath83 where @xmath84 denotes the @xmath0-pochhammer symbol . as before , we use creative telescoping to provide a certified operator @xmath85 which annihilates the sum on the left - hand side .
in the present case , a single operator is sufficient because the sum depends only on a single variable @xmath18 ( there is no @xmath86 there ) . using the operator @xmath85 and the defining equations for @xmath24
, we then construct a recurrence for the entire left - hand side , which turns out to have order twelve . as
this recurrence is a left multiple of the second order operator annihilating the right - hand side , the proof of theorem [ qtsppthm ] is completed by checking for @xmath87 .
_ quod erat demonstrandum _ !
paul erds famously ( and naively ) believed that every short and elegant mathematical statement has a short and elegant `` proof from the book '' , and if humans tried hard enough , they would eventually find it .
kurt gdel , on the other hand , meta - proved that there exist many short statements whose shortest possible proof is very long .
it is very possible that the @xmath0-tspp theorem _ does _ have a yet - to - be - found _ proof from the book _ , but it is just as possible that it does not , and while we are sure that the present proof is not the shortest - possible ( there were lots of random choices in designing the proof ) , it may well be the case that the shortest - possible proof is still very long , and would still require heavy - duty computer calculations . be that as it may , we believe that our general _ approach _ , and our way of _ taming _ the computer to prove something that seemed intractable with today s hardware and software are very elegant , and deserve to be included in the _ book_. richard stanley , _ a baker s dozen of conjectures concerning plane partitions_. in _ combinatoire numrative _ , ed .
g. labelle and p. leroux .
lecture notes in mathematics * 1234 * , 285293 .
springer verlag , new york . | the conjecture that the orbit - counting generating function for totally symmetric plane partitions can be written as an explicit product - formula , has been stated independently by george andrews and david robbins around 1983 .
we present a proof of this long - standing conjecture . supported by nfs - dms 0070567 as a postdoctoral fellow and by the austrian science foundation ( fwf ) grant y464-n18 . ]
corresponding author . supported by the austrian science foundation ( fwf ) grant y464-n18 . ] supported in part by the united states of america national science foundation . ] |
the mass distribution of stars at birth , known as the initial mass function ( imf ) , is perhaps the most important fundamental parameterization of the star formation process ( e.g. , lada 2009 ; mckee & ostriker 2007 ; adams 2003 ) .
when salpeter ( 1955 ) first empirically evaluated the imf , he identified a simple power - law distribution for stellar mass @xmath9 , having exponent @xmath10 : @xmath11 while it is now clear that the imf has a characteristic mass below which its form turns over , yielding additional distinct regimes at lower masses ( e.g. , kroupa 2001 ; chabrier 2003 ) , salpeter s original prescription nevertheless has remained surprisingly robust for @xmath12 . since the stellar light from galaxies is dominated by these more luminous stars , the `` salpeter imf '' has proven vital to the entire field of galaxy evolution , in addition to star formation and stellar populations . yet , more than half a century after salpeter s ( 1955 ) seminal work , the physical factors yielding the value of @xmath13 for the power - law index remain elusive and poorly understood .
the origin of the imf continues to be a topic of intense discussion , and the reader is referred to recent reviews by lada ( 2009 ) , clarke ( 2009 ) , elmegreen ( 2009 ) , mckee & ostriker ( 2007 ) , bonnell , larson & zinnecker ( 2007 ) for comprehensive overviews on imf theory .
one obstacle to gaining deeper physical understanding is the fact that the salpeter slope emerges from a variety of simulations dominated by different mechanisms . in particular , both the core collapse ( e.g. , krumholz et al . 2010 ) and competitive accretion scenarios ( e.g. , bonnell et al .
2003 ) are able to reproduce the salpeter slope .
effects that are explored in modern simulations include decaying vs driven turbulence , isothermal vs non - isothermal equations of state , and inclusion or exclusion of magnetic fields , among other factors ; the imf slope generally does not provide a strong discriminant on this large parameter space ( e.g. , clarke 2009 ; mckee & ostriker 2007 ) . here
, we suggest an explanation of the salpeter slope that may provide the simplest basic physical understanding , but which may correspond to a framework within which the more specific physics of actual star formation operates .
most , if not all , stars form in clusters ( e.g. , mckee & ostriker 2007 ; lada & lada 2003 ) , which in turn form with an initial cluster mass function ( icmf ) . like the stellar imf
, the cluster icmf is also described well by a power - law distribution in cluster mass @xmath14 : @xmath15 again like the stellar imf , the icmf power law index is also found to be robust and largely invariant , but slightly shallower , @xmath16 , than the salpeter slope , based on clusters covering a wide range of scales , including super star clusters ( e.g. , meurer et al .
1995 ) , massive and open clusters ( e.g. , elmegreen & efremov 1997 ; zhang & fall 1999 ; de grijs et al .
2003 ) , sparse ob groups ( oey et al .
2004 ) , and the hii region luminosity function ( e.g. , oey & clarke 1998 ; kennicutt et al .
1989 ) .
the similarity in the slopes of the imf and icmf , @xmath10 vs @xmath17 , respectively , has been previously noted ( e.g. , elmegreen 2006 ) . since star formation is a hierarchical process , with smaller units fragmenting from larger ones , we can examine the origin of the imf as the relationship between the imf and icmf .
the icmf slope @xmath18 is a value that seems simpler to understand .
as is often pointed out , this power - law index corresponds to a uniform distribution of power between large and small scales ( e.g. , elmegreen 2006 ) .
furthermore , zinnecker ( 1982 ) showed that an imf slope @xmath19 results simply from bondi - hoyle accretion @xmath20 among the protostellar masses , with the mass function @xmath21 evolving from some arbitrary initial distribution : @xmath22 this yields a simple power law imf having @xmath0 ( see also bonnell et al .
2007 ; hsu et al . 2010 ) .
thus @xmath0 is a value that is intuitive , with simple and reasonable physical bases from both the parent icmf equipartition arguments and build - up from simple accretion .
so then why is the salpeter slope of the imf slightly steeper than @xmath0 ?
elmegreen ( 2009 ) summarizes the three approaches to generating the imf : 1 ) fragmentation , in which the core mass function and imf is a direct mapping from the parent cloud fragmentation ; 2 ) accretion , in which the imf is the product of protostar accretion processes not necessarily linked to how clouds fragment ; and 3 ) interruption of 1 ) or 2 ) , in which the form of the imf is also determined by factors that preferentially limit fragmentation or accretion for high or low - mass stars .
if we adopt our argument above that , simplistically , the imf should inherit @xmath23 from the icmf according to both 1 ) and 2 ) , then 3 ) is a likely candidate to explain the deviation of the salpeter slope : there is likely some process that limits the formation of high - mass stars in favor of low - mass ones . in a fully hierarchical star formation scenario , the masses of the parent clouds , and
hence , cluster masses , are determined before the masses of the star - forming clumps , and in turn , the constituent stars .
if the smallest clumps have masses on the order of stellar masses , then the formation of high - mass stars will necessarily be limited in such clumps , relative to the low - mass stars . in recent years
, it has been proposed that the integrated galaxy imf ( igimf ) has a steeper slope than the imf because the most massive stars may not form in smallest clusters ( kroupa & weidner 2003 ; weidner & kroupa 2005 , 2006 ; see also elmegreen 1999 ) , causing the maximum stellar mass @xmath24 to depend on the cluster mass .
this would cause the igimf to favor lower - mass stars .
we propose that the difference between the salpeter slope @xmath25 and the icmf slope @xmath18 is caused by a similar situation governing the distribution of stellar masses within individual clusters . if , following a fully hierarchical scenario , stellar subgroups are formed out of clumps with pre - determined masses , and if the clump mass distribution also generates a dependence of @xmath24 on clump mass @xmath14 , then the resulting imf for the entire cluster will be slightly steeper than the clump mass function .
this scenario has essentially been proposed by elmegreen ( 1997 ) , who presented this as random sampling from a fractal mass distribution .
here , we present a parameterization that is much simpler than elmegreen s model , and we show that _ the salpeter slope simply results from the overlap between the stellar mass range and the clump mass range . _
we model the generation of stars in a cluster as the sum of all stars forming from individual clumps in the parent molecular cloud .
we assume that the hierarchical fragmentation of the cloud into clumps follows the same @xmath18 power - law distribution found for the icmf , and further , that the fragmentation of clumps into stars does the same , arguing that self - similar physical mechanisms govern all these fragmentation processes on smaller scales .
this breaks down around the characteristic value of @xmath3 , suggested to be linked to the jeans mass ( e.g. , mckee & ostriker 2007 ; clarke 2009 ) .
since we are interested only in the upper imf , however , scale - free cloud fragmentation plausibly dominates in this regime .
but we note that our model contrasts with those in which the imf slope results from physics that causes fragmentation into a steeper core mass distribution ( e.g. , padoan & nordlund 2002 ) .
we construct monte carlo simulations drawing the clump mass function from the @xmath18 power law , and then the stellar mass function within each clump from the same distribution , within lower and upper stellar mass limits @xmath26 and @xmath24 .
the entire cluster is then the aggregate of all the stars formed in all the clumps , and the imf is the composite for this aggregate .
our default sampling algorithm allows rejection of stars that cause the clump mass limit to be exceeded , and continues sampling until the total stellar mass is within @xmath27 of the specified clump mass .
this is similar to , but a bit more strongly sorted , than the `` sorted sampling '' algorithm of weidner & kroupa ( 2006 ; see also elmegreen 2006 ; parker & goodwin 2007 ) . on the allowed mass range for stars ( @xmath27 to @xmath24 ) relative to that for clumps ( @xmath28 to @xmath29 ) .
panels @xmath30
@xmath31 are based on sorted sampling , and panel @xmath32 is based on unsorted sampling . to ease comparison
, the black symbol shows the same model parameters , and the dotted line shows the salpeter slope value.,width=264 ] figure [ f_slopemodels ] shows how the observed logarithmic imf slope @xmath33 , determined from fits weighted by the inverse of the poisson errors , depends on the mass ranges for the stars relative to the clumps .
if the clump masses are all larger than the allowed stellar masses , then the stellar masses are unconstrained , and the parent @xmath18 distribution is reproduced , translating to @xmath34 on a logarithmic plot .
this is apparent in figure [ f_slopemodels]@xmath30 , having a fixed allowed stellar mass range of 1 to 100 @xmath35 , and clump upper - mass limit @xmath36 .
figure [ f_slopemodels]@xmath30 shows @xmath33 as a function of the lower clump limit @xmath28 , and as @xmath37 , @xmath38 . _ but as allowed cloud masses decrease to values below @xmath24 , the imf slope steepens , since formation of the most massive stars is excluded .
_ we see that as @xmath39 , the value approaches @xmath40 , essentially the salpeter value ( black point ) .
figure [ f_slopemodels]@xmath41 confirms that this effect is driven by the allowed clump lower - mass limit , rather than the upper limits on either the clumps or stars .
this model allows both @xmath29 and @xmath24 to be essentially limitless , @xmath42 ; the results are almost identical to figure [ f_slopemodels]@xmath30 in approaching the same salpeter slope as @xmath39 .
this steepening effect between @xmath18 and the salpeter value can be seen in models by elmegreen ( 1997 ) and weidner & kroupa ( 2006 ) .
indeed , elmegreen s model allowed an effective equivalent @xmath43 , and weidner & kroupa adopted an analogous @xmath44 ; both of these clump lower - mass limits are well within the regime where the imf steepening is expected , according to our simulations .
imf slopes steeper than the salpeter value can be obtained if _ both _ @xmath45 and @xmath29 approaches stellar values .
figure [ f_slopemodels]@xmath31 assumes @xmath46 , and shows @xmath33 as a function of @xmath29 . for @xmath47 , @xmath48 , as found before for the assumed @xmath49 . as the clump upper - mass limit
is decreased , @xmath33 steepens , reaching @xmath50 for @xmath51 .
the results are essentially identical even if @xmath24 is limitless , although the maximum stellar mass will be limited by @xmath29 at low @xmath14 .
weidner & kroupa ( 2006 ) also show that the degree to which the aggregate imf steepens depends slightly on the sampling algorithm used .
a `` sorted sampling '' algorithm , as we use , allows continued sampling of lower - mass stars after the highest - mass stars can no longer fit within the alloted mass , and naturally induces slightly steeper slopes than algorithms that stop sampling as soon as any drawn star causes the clump mass limit to be reached .
figure [ f_slopemodels]@xmath32 is the same as panel @xmath31 , but with unsorted sampling , confirming that the sorting scheme is not a strong effect .
thus we see that the slope steepening beyond the input @xmath18 value is driven by allowing clump masses @xmath52 . in these lowest - mass clumps ,
the highest - mass stars can not form , thus causing the aggregate cluster to slightly favor lower - mass stars beyond the conventional imf . from a physical standpoint ,
this simply means that clumps too small to produce the highest - mass stars can still produce low - mass stars .
furthermore , the salpeter imf results when @xmath53 while randomly sampling both stars and parent clumps from the same @xmath18 power - law distribution .
thus ironically , the critical parameter for steepening the imf and the igimf is not @xmath29 , but rather @xmath28 .
elmegreen ( 1997 , 1999 ) developed a model for imf generation based on random sampling of stellar masses from hierarchical fractal structure .
our model distills the effects seen in elmegreen s model to the most simplististic level , and it more directly captures the dominant effect seen in that work , revealing @xmath28 as the critical parameter .
the essential condition that drives the steepening of the imf from @xmath23 is that @xmath45 . this need not be interpreted literally , but the smallest clumps must be incapable of producing the highest - mass stars ; our model is independent of star - formation efficiency , provided that it is essentially constant . to attain the salpeter value , in particular , requires the equivalent of @xmath53 .
because the full imf flattens strongly near a characteristic value @xmath54 , this limits the power - law form of the high - mass imf to apply only above this characteristic mass , thus @xmath55 , recalling that here we consider only the upper , salpeter imf .
furthermore , because stellar objects continue to form with masses 1 2 orders of magnitude lower , clump fragmentation clearly continues to masses below @xmath56 as well , and so we know that the relevant lowest - mass clumps having @xmath55 do form stars . thus , for the power - law upper imf , @xmath57 , which is exactly the condition needed to induce steepening of the aggregate cluster imf to the salpeter value in the power - law regime above @xmath56 .
in other words , @xmath56 acts as an effective lower cutoff to the imf power - law distribution for both clump and stellar masses , rendering @xmath58 . indeed ,
c@xmath59o observations of m17-sw show that molecular clump masses follow a power law of @xmath60 down to at least 1 @xmath35 ( stutzki & gsten 1990 ) , while the observed imf in the closely associated region m17-swex shows @xmath61 ( povich & whitney 2010 ) .
if clumps fragment according to a @xmath18 power law , then the range of observed imf slopes should generally not fall below this value .
similarly , steeper imf slopes reach @xmath62 for @xmath51 , when @xmath53 . a number of authors have shown compilations of measured cluster slopes , and @xmath63 does appear to be a fairly well - defined range of allowed slopes for the high - mass imf ( e.g. , elmegreen 1999 , fig . 1 ; bastian et al .
2010 , fig .
2 ) : what appears to be a large scatter about @xmath1 ( elmegreen 1999 ) can also be interpreted as variation within the range allowed by our model .
the basic premise of our model is that clumps fragment from clouds according to the @xmath18 power law , and in particular , that stars formed within each individual clump are still formed with a @xmath0 power law .
thus if newly - formed stars can be identified with their natal sub - groups in the youngest star - forming regions , the imf should appear to be closer to @xmath64 in such groups .
it would be interesting to evaluate the imf slopes for sub - groups within massive star - forming regions , and to compare these to the aggregate cluster imf .
also , observations focusing on specific subregions within extremely young clusters might be more likely to obtain imf slopes closer to @xmath64 , while more complete measurements for entire clusters might obtain the salpeter value or higher .
these are difficult tests , since dynamical evolution will quickly mix stars born from different clumps .
it is often suggested ( e.g. , elmegreen 2004 , 2009 ) that the imf is slightly flatter in starbursts and rich , young , super star clusters . in our model , recovering an imf slope of @xmath0 occurs naturally if the clump mass @xmath65 .
the arches cluster near the galactic center is the best - known example of a flat imf , with the most recent measurements of the high - mass imf of @xmath66 ( kim et al .
2006 ) and @xmath67 ( espinoza et al .
2009 ) ; these values are consistent with our model .
regions of extreme star formation are expected to have higher thermal jeans masses owing to radiative feedback from massive stars ( e.g. , larson 2005 ; murray 2009 ) therefore raising the minimum clump mass @xmath28 .
our slope flattening would only occur if the clump mass range is increased without a commensurate increase in the allowed stellar mass range ; and indeed , @xmath24 appears to be remarkably constant and independent of star formation environment ( e.g. , oey & clarke 2005 ; weidner & kroupa 2004 ) . if the smallest clusters form out of single clumps , with no further fragmentation , then the cluster imf is the same as the clump imf and there is no steepening of the aggregate slope .
hence , such single - clump clusters should show the parent @xmath0 stellar imf .
further , if the smallest clusters do approach limiting masses that are on the order of the largest clump masses , this should cause an analogous steepening of the cluster mass function under these circumstances . as described above
, the icmf generally has a slope @xmath18 , but it would be interesting to evaluate any trends .
we also note that our model does not necessarily imply an igimf steepening itself : smaller clusters can still fragment into clumps yielding the full stellar mass range for all , including the smallest , clumps .
so if the smallest clusters are still capable of forming stars up to masses @xmath24 , as suggested by the results of lamb et al .
( 2010 ) , then no igimf steepening will occur . thus again
, the lower clump mass range relative to the upper stellar mass range determines the steepening for both the cluster stellar imf and the igimf .
_ we do however suggest that the igimf effect seen in low - density environments _ ( e.g. , lee et al . 2009 ; hoversten & glazebrook 2008 ) _ is real and occurs because the smallest clusters have masses below @xmath24 .
_ we stress that our model is a simplistic parameterization of hierarchical star formation , but it may capture essential elements that govern the generation of the salpeter tail of the imf . in particular , our model is based on a scenario that is fully hierarchical , with clouds fragmenting into clumps and clumps into stars , with each fragmentation process based on a @xmath68 power - law mass distribution .
`` fragmentation '' need not be literal , but simply means that the ultimate mass apportionment from the parent to descendant units follows this power - law distribution , whether by competitive accretion or by well - defined fragmentation .
the steepening of the imf to the salpeter value is then induced by the inability of the lowest - mass clumps to form the highest - mass stars .
maschberger et al .
( 2010 ) have carried out a star - formation simulation of a large volume , which follows the hierarchical fragmentation of ism into clusters and down to core - like sink particles .
their results are consistent with our scenario , showing that clusters are mergers of products from multiple clumps , and that the aggregate imf in such mergers is slightly steeper than in the individual merging units .
we propose that the empirical , logarithmic @xmath1 salpeter power - law slope of the imf for high - mass stars originates from a universal , linear @xmath0 imf slope for stellar sub - groups formed from clumps within individual clusters .
our monte carlo simulations show that @xmath69 is the limiting value of the imf slope when the lower - mass limit for both clumps and stars is equal , @xmath49 ; this condition holds for the high - mass tail of the observed imf , since it turns over below the characteristic mass @xmath3 , effectively setting the lower - mass limit to both @xmath27 and @xmath28 .
the steepening to the salpeter value occurs because the highest - mass stars can not form in the lowest - mass clumps .
thus , a critical factor to examine in star formation theory is the effective mass range for clumps relative to the output stellar masses .
our model is analogous to the mechanism for steepening the igimf proposed by kroupa & weidner ( 2003 ) , but a universal salpeter slope does not necessarily imply that the igimf should steepen to even higher values , since the smallest clusters still may be capable of forming the highest - mass stars .
however , we stress that the same principles involving the relative mass ranges of the parent and descendant units also applies to the igimf and may be responsible for the apparent steepening in regimes dominated by the smallest - scale star formation ( e.g. , lee et al .
2009 ) .
our model is based on a fully hierarchical scenario in which the physics of fragmentation from large scales to individual stars takes place in a self - similar manner , all with a 2 power - law mass distribution .
thus , this does not require literal fragmentation , but simply that the final mass distribution among both the parent and descendant units is distributed accordingly . with this simple condition
, we show that the aggregate high - mass imf slope is limited between values of @xmath64 and 2 , which is consistent with observations .
flatter values near @xmath64 occur when @xmath8 , which may explain the suggested imf flattening for regions of extremely intense , massive star formation . while our model is only a rough parameterization of the star formation process , we suggest that it may capture the fundamental effects that generate the salpeter slope of imf for high - mass stars .
thanks to cathie clarke , joel lamb , and fred adams for useful discussions and comments , and to the referee , michael reid .
i gratefully acknowledge the institute of astronomy , cambridge for visitor support , and funding by nsf grant ast-0907758 .
adams , f. c. 2003 , pub .
beograd , 75 , 9 bastian , n. , covey , k. r. , & meyer , m. r. 2010 , ara&a , 48 , 339 bonnell , i. a. , bate , m. r. , & vine , s. g. 2003 , mnras 343 , 413 bonnell , i. a. , larson , r. b. , & zinnecker , h. 2007 , protostars and planets v , 149 chabrier , g. 2003 , pasp 115 , 763 clark , c. j. 2009 , apss 324 , 121 de grijs , r. , andres , p. , bastian , n. , lynds , r. , lamers , h. j. g. l. m. , & oneil , e. j. 2003 , mnras , 343 , 1285 elmegreen , b. g. 1997 , apj 486 , 944 elmegreen , b. g. 1999 , apj 515 , 323 elmegreen , b. g. 2002 , apj 564 , 773 elmegreen , b. g. 2004 , apj 354 , 367 elmegreen , b. g. 2006 , apj 648 , 572 elmegreen , b. g. 2009 , in _ the evolving ism in the milky way and nearby galaxies , _ k. sheth , a. noriega - crespo , j. ingalls , and r. paladini , http://arxiv.org/abs/0803.3154 espinoza , p. , selman , f. j. , & melnick , j. 2009 , a&a , 501 , 563 hoversten , e. a. & glazebrook , k. 2008 , apj 675 , 163 hsu , w .- h . , hartmann , l. w. , heitsch , f. , & g mez , g. c. 2010 , apj 721 , 1531 .
kennicutt , r. c. , edgar , b. k. , & hodge , p. w. 1989 , apj 337 , 761 kim , s. s. , figer , d. f. , kudritzki , r. p. , & najarro , f. 2006 , apjl , 653 , l113 kroupa , p. & weidner , c. 2003 , apj 598 , 1076 krumholz , m. r. , cunningham , a. j. , klein , r. i. , & mckee , c. f. 2010 , apj 713 , 1120 lada , c. j. 2009 , in _ astrophysics in the next decade , _ h. a. thronson , m. stiavelli , a. tielens , springer netherlands , 147 lada , c. j. & lada , e. a. 2003 , ara&a , 41 , 57 lamb , j. b. , oey , m. s. , werk , j. k. , & ingleby , l. d. 2010 , apj 725 , 1886 larson , r. b. 2005 , mnras 359 , 211 lee , j. c. , et al .
2011 , apjs , 192 , 6 maschberger , t. , clarke , c. j. , bonnell , i. a. , & kroupa , p. 2010 , mnras , 404 , 1061 mckee , c. f. & ostriker , e. c. 2007 , ara&a 45 , 565 meurer , g. r. , heckman , t. m. , leitherer , c. , kinney , a. , robert , c. , & garnett , d. r. 1995 , aj 110 , 2665 murray , n. 2009 , apj 691 , 946 oey , m. s. & clarke , c. j. 1998 , aj 115 , 1543 oey , m. s. & clarke , c. j. 2005 , apjl 620 , l43 oey , m. s. , king , n. l. , & parker , j. w. 2004 , aj 127 , 1632 padoan , p. & nordlund , 2002 , apj 576 , 870 parker , r. j. & goodwin , s. p. 2007
, mnras 380 , 1271 povich , m. s. & whitney , b. a. 2010 , apj 714 , l285 salpeter , e. e. 1955 , apj 121 , 161 stutzki , j. & gsten , r. 1990 , apj 356 , 513 weidner , c. & kroupa , p. 2004 , mnras 348 , 187 weidner , c. & kroupa , p. 2005
, apj 625 , 754 weidner , c. & kroupa , p. 2006 , mnras 365 , 1333 zhang , q. & fall , s. m. 1999 , apjl 527 , l81 zinnecker , h. 1982 , annals ny acad . | we suggest that the intrinsic , stellar initial mass function ( imf ) follows a power - law slope @xmath0 , inherited from hierarchical fragmentation of molecular clouds into clumps and clumps into stars .
the well - known , logarithmic salpeter slope @xmath1 in clusters is then the aggregate slope for all the star - forming clumps contributing to an individual cluster , and it is steeper than the intrinsic slope within individual clumps because the smallest star - forming clumps contributing to any given cluster are unable to form the highest - mass stars .
our monte carlo simulations demonstrate that the salpeter power - law index is the limiting value obtained for the cluster imf when the lower - mass limits for allowed stellar masses and star - forming clumps are effectively equal , @xmath2 .
this condition indeed is imposed for the high - mass imf tail by the turn - over at the characteristic value @xmath3 .
imf slopes of @xmath4 are obtained if the stellar and clump upper - mass limits are also equal @xmath5 , and so our model explains the observed range of imf slopes between @xmath6 to 2 .
flatter slopes of @xmath7 are expected when @xmath8 , which is a plausible condition in starbursts , where such slopes are suggested to occur .
while this model is a simplistic parameterization of the star - formation process , it seems likely to capture the essential elements that generate the salpeter tail of the imf for massive stars .
these principles also likely explain the igimf effect seen in low - density star - forming environments . |
the theoretical description of the fusion of nuclei has so far relied on a many - degrees - of - freedoms tunneling formalism referred to as the coupled channels theory . in the fusion of stable nuclei
the channels considered correspond to bound states in the participating nuclei .
when weakly bound nuclei fuse , the coupling to continuum channels becomes important and one resorts to by and large intuitive means to take these into account @xcite . if fusion occurs above the coulomb barrier , quantum tunneling is less important and one uses geometrical arguments to consider these couplings .
in fact this is also done in the description of the fusion of fullerene molecules @xcite , important in the development of strong nano - tubes . at near or sub - barrier energies
, tunneling becomes a dominant feature of the fusion of nuclei and considering the coupling to the continuum or breakup channels has been a challenge to theorists .
the reason being the need for a full three - body model for tunneling that describes both the complete fusion as well as the incomplete fusion of one of the fragments when breakup takes place . in this three - body model
the interactions have to be complex .
so far no such complete theory is practically developed , although serious attempts towards this goal have been made @xcite .
less ambitious endeavours based on a discretized continuum within a basically extended two - body model have been extensively used .
this is the continuum discretized coupled channels ( cdcc ) method .
variants of this method have also been pursued especially in the fusion of radioactive nuclei . in this contribution
we present a short account of some of these theoretical attempts to treat the quantum tunneling of fragile objects .
let us first consider the fusion cross section in potential scattering , where the components of the projectile - target separation vector are the only degrees of freedom taken into account .
the collision dynamics is described by the hamiltonian @xmath0 where @xmath1 is the kinetic energy operator and @xmath2 is the optical potential . in typical situations
, one has @xmath3 where @xmath4 and @xmath5 are absorptive potentials , accounting for the flux lost to fusion and to peripheral processes , respectively .
the scattering state @xmath6 is obtained solving the schrdinger equation with scattering boundary condition and the fusion cross section is @xmath7 above , @xmath8 is the collision energy and @xmath9 is the wave number associated to it . in multi - channel scattering the problem is more complicated since the hamiltonian depends also on intrinsic degrees of freedom .
the scattering wave function is expanded in eigenstates of the intrinsic hamiltonian , @xmath10 with eigenvalues @xmath11 , in the form @xmath12 and the channel wave functions , @xmath13 , are given by the coupled - channel equations , @xmath14 ~\psi_{\alpha}^{(+ ) } = \sum_{\beta\left ( \neq\alpha\right ) } u_{\alpha,\beta}~\psi_{\beta } ^{(+)};\;\;\;\;\;\alpha,\beta=0,1, .... ~. \label{cc - eq}\ ] ] above , @xmath15 , @xmath16 if all relevant direct channels are included in the expansion , the imaginary potential @xmath17 represents exclusively fusion absorption .
if the off - diagonal part of the interaction , @xmath18 is real , as frequently is the case , the fusion cross section can be written @xmath19 when the channel coupling interaction is complex , off - diagonal terms should also be included in the summation above .
it is well known that channel coupling leads to a strong enhancement of the fusion cross section at sub - barrier energies @xcite .
that is , the sum of the contributions from all channels is larger than the fusion cross section when the coupling is switched off .
however , it is important to investigate the effect of the coupling on the fusion through the elastic channel alone , @xmath20 for this purpose it is convenient to introduce the polarization potential .
although its calculation may be difficult , it provides a convenient language for the discussion .
this potential is defined by the condition that the same elastic wave function @xmath21 of the coupled - channel equations be obtained from a potential scattering equation ( eq .
( [ cc - eq ] ) with the rhs set equal to 0 ) , with the replacements @xmath22 the inclusion of @xmath23 always reduces the fusion cross section , since it leads to absorption of the incident flux before it reaches the region where fusion takes place .
the real part of the potential may lead to additional fusion suppression , if it is repulsive .
however , if it is attractive , it reduces the barrier height and thus enhances the fusion cross section . in this case
, there is a competition between the real and the imaginary parts of the polarization potential and the net result depends on the details of that potential .
collisions of weakly bound nuclei are much more complicated since the elastic channel couples strongly with continuum states . in this case , the channel label becomes continuous and there is an infinite number of coupled channels , even with truncation at low values of the excitation energy . a theory including continuum coupling and describing the various fusion processes is very hard to develop .
an important feature of these collisions is that the breakup mechanism may give rise to two different types of fusion .
complete fusion ( cf ) , when the whole masses of the projectile and the target are contained in the compound nucleus , and incomplete fusion ( icf ) , when some nucleons move out of the interaction region before the formation of the compound nucleus .
an additional complication is that cf can take place in a single step or in a sequential way , as the fragments of the broken projectile are successively absorbed by the target .
these mechanisms are called direct ( dcf ) and sequential ( scf ) complete fusion , respectively .
the sum of cf and icf is termed total fusion ( tf ) . in this section ,
we summarize some of the attempts available in the literature , with emphasis on the continuum discretized coupled channels method ( cdcc ) .
the earliest calculation of fusion cross sections in collisions of weakly bound nuclei were carried out with schematic models , based on drastic approximations .
the static effects of the halo could easily be taken into account through the use of larger diffusivities in the real potential . however , the inclusion of dynamic effects is a much harder task , since it involves the coupling with continuum states . in the first calculations , coupled channel effects
were mocked up by an approximate polarization potential .
since the adopted approximations led to a purely imaginary potential , changes in the barrier hight were simulated by a shift in the collision energy .
these works found that the cf cross section was hindered around the barrier energy @xcite .
soon after these works , a calculation adopting the opposite view was performed by dasso and vitturi @xcite .
these authors performed coupled channel calculations in a simple model in which the continuum was represented by a bound effective channel .
the cf cross section was then obtained from eq .
( [ sigfus ] ) , with the exclusion of the effective channel from the summation . however , in this model important properties of the continuum were not properly described . in opposition to the previous results
, they obtained a strongly enhanced cf cross section . since these calculations were based on very schematic models , using different approximations , it is not surprising that they led to conflicting conclusions .
it is clear that a reliable conclusion about the dynamic effects of the coupling requires the use of more realistic models .
the cdcc method consists of replacing the continuous part of the projectile s spectrum by a finite number of discrete states .
this can be done in different ways @xcite .
the most frequent approach is to divide the continuum in bins , distributed up to some cut - off energy @xmath24 , and to build wave packets within each bin .
the wave packets are superpositions of energy eigenstates @xmath25 , where @xmath9 is the wave number associated with the relative motion of the projectile s fragments and @xmath26 stands for the remaining projectile s quantum numbers .
one common option is to use @xmath27 bins of equal size , @xmath28 , in momentum space , distributed between 0 and @xmath29 . in this case
, the @xmath30 discrete state with quantum numbers @xmath26 is @xmath31 the particular shapes of the wave packets do not significantly affect the results .
however , they should be chosen such that the states @xmath32 form an orthonormal set . in this way
the problem is reduced to a set of coupled - channel equations involving the elastic , the excited bound channels and @xmath27 discretized channels for each @xmath26 .
an example of the application of this method to the @xmath33b breakup process may be found in ref .
the comparison of the full cdcc calculations with other simplified calculations shows the importance of higher - order couplings , and also of the couplings between continuum states .
this last ingredient , continuum - continuum couplings , had not been included in the calculations of the fusion cross section for the @xmath34be + @xmath35 pb collision presented in ref .
their results are shown in figure [ fig1 ] , in comparison with predictions of a bare potential ( channel - coupling switched off ) .
within those conditions , the cf cross section is strongly enhanced at low energies and slightly hindered at high energies .
the transition between these behaviors occurs at an energy @xmath36 .
thus , this study leads to enhancement of the cf cross section in the barrier region .
the effect of the continuum - continuum couplings on this reaction was investigated by diaz - torres and thompson @xcite .
their results are shown in figure [ fig2 ] .
continuum - continuum couplings were not included in the cdcc calculations presented in ( a ) and were included in the ones shown in ( b ) . in ( a ) , the results are similar to those of ref .
@xcite while in ( b ) they are rather different .
both the cf and tf cross sections are much smaller than those obtained without continuum - continuum couplings . in this case , the cf cross section is suppressed in the barrier region , with @xmath37 falling below the coulomb barrier .
it should be remarked that neither of these works take into account the sequential capture of all fragments , and thus they provide only a lower limit for the cf cross section .
intuitively , the reduction of the fusion cross section arising from couplings among many channels in the continuum sounds reasonable .
it is , however , of interest to prove this formally .
one can formulate the problem using feshbach theory and find a way of eliminating the continuum - continuum coupling terms .
what one finds is a set of coupled channels equations of the type : @xmath38 where @xmath39 and @xmath40 are different operators .
now we know from compound nucleus theory that if this happens ( the c - channels are closed here ) then time reversal is violated according to the definition of wigner of time reversal invariance @xcite .
if time reversal is violated one has irreversibility .
however , what we are interested here is the so - called statistical mechanics irreversibility . of course ,
in order for this to happen , one has to resort to coarse graining the continuum - continuum couplings ( energy - averaging ) as it has been done extensively in the past in the treatment of deep inelastic collisions @xcite .
this may indeed be happening in the actual calculations using the cdcc method .
new calculations employing the cdcc method to determine the effect of the breakup channel on the total fusion cross section in reactions induced by @xmath41li projectiles were performed more recently @xcite . the main difference between these calculations and those of ref .
@xcite is in the absorptive interaction .
@xcite uses a projectile - target imaginary potential while ref .
@xcite considers absorption of each projectile s fragment separately . in this way
, only the tf cross section is evaluated .
it was found through those calculations @xcite that the presence of the breakup channel enhances the total ( complete + incomplete ) fusion cross section in the region around the coulomb barrier , and it has very little effect at higher energies .
it was further observed that the effect depended on the target size , as expected , being more marked in the case of a @xmath42bi than for @xmath43co .
the differences in total fusion cross section between the two lithium projectiles may be attributed to the coupling to the breakup channel in the case of @xmath43co , but not for the @xmath42bi target .
one should also mention the combined cdcc - dynamic polarization potential ( dpp ) approach introduced by keeley , rusek and collaborators @xcite .
the procedure consists in first performing cdcc calculations to describe the elastic scattering data for the system under consideration , and from the calculations extract the corresponding polarization potentials .
the fusion cross sections are then obtained using a simple barrier penetration calculation , where the barrier is given by the addition of the real parts of the optical potential and the polarization potential .
this cdcc - dpp method was employed to extract the effect of the @xmath44he dipole polarizability on its elastic scattering on @xmath35pb , and its correspondent effect on the fusion cross section for this reaction .
a comparison of the theoretical fusion cross section of the projectiles @xmath44he and @xmath44li incident on @xmath35pb is shown in fig.[fig3 ] . as there is no available data for these systems ,
the calculations are compared with data for the @xmath44he on @xmath42bi fusion cross section .
the figure shows results for these projectiles with and without coupling to the breakup channel . when this coupling is left out ( only static effects ) , the cross section for @xmath44he is much larger than that for @xmath44li , owing the @xmath44he neutron halo .
when breakup couplings are taken into account , repulsive potentials reduce the cross sections .
since this effect is stronger for @xmath44he , the static effect of the halo is cancelled and the fusion cross section for the two projectiles are similar . the cdcc cross section for @xmath44he projectiles is consistent with the data .
he + @xmath45pb and @xmath46li + @xmath35pb systems .
data for @xmath46he + @xmath47bi are also shown ( figure from ref .
@xcite).,width=377 ] the breakup of the @xmath44he and @xmath44li projectiles proceeds according to different mechanisms . in the case of @xmath46he ,
the @xmath81 transitions to the continuum dominate .
as the @xmath48he and @xmath49h fragments of @xmath44li have the same charge / mass ratios , these transitions do not play a role in the case of @xmath44li , which breakup is thus produced by quadrupole couplings , mainly driven by nuclear forces .
besides the quantal coupled - channels procedures mentioned above , other alternative methods have been employed to describe the fusion and breakup processes in reactions induced by weakly bound projectiles .
recently , hagino and collaborators have developed a classical trajectory method , in which the classical trajectories for the target and the two projectile fragments are evaluated for a monte carlo sampling of initial conditions @xcite .
this method automatically separates the complete and incomplete fusion cross sections , so it does not suffer from the deficiency of the cdcc calculations presented in @xcite which count the sequential fusion of all fragments as an incomplete , rather than a complete fusion process .
in fact , according to these classical calculations , the cross section for the capture of all fragments represents an important contribution to the complete fusion cross section .
the results of these calculations are in reasonable agreement with the experimental data at energies above the coulomb barrier .
however , as they do not include tunneling effects , they can not be employed at the underbarrier energies where much of the interest in fusion induced by weakly bound projectiles lies .
a time - dependent description of the collision process was developed to study fusion reactions induced by halo nuclei @xcite . in this approach
the system is described as a three - body system , composed of the target , the projectile core and the halo nucleon .
this three - body reaction problem is described in terms of the time evolution of a wave packet .
the initial wave packet is taken to be the product of an incoming wave for the relative motion between the centers of mass of the projectile and target nuclei , multiplied by a gaussian wave packet representing the initial configuration of the projectile fragments .
the evolution of this wave packet is treated through the standard time - dependent schrdinger equation approach , which although requiring long calculation times , is simple to program .
the fusion probability is calculated through an energy projection procedure .
this approach has been applied to several systems @xcite . in the case of neutron halos , it is found that the effect of the halo hinders the fusion cross section . since semiclassical calculations of continuum discretization coupled - channels have been successfully applied to the description of the breakup of weakly bound projectiles @xcite , they should be extendable to the treatment of fusion reactions .
the breakup amplitudes calculated with the semiclassical calculations would determine the initial conditions for the study of the fragments .
such a procedure would have the advantage of automatically separating the motion of all fragments , thus separating the sequential fusion contribution to complete fusion from the incomplete fusion process , which was one of the difficulties facing the quantum mechanical cdcc calculations .
a preliminay calculation of the cf cross section in a schematic two - channel model has recently been performed @xcite and the results were compared with results of the coupled channel method .
the agreement was good .
in this paper we have presented a brief review of the theoretical methods to describe fusion reactions with weakly bound nuclei .
we pointed out that these reactions involve the coupling to continuum states , which is very hard to describe . in this way
, simple models using drastic approximations can hardly provide a reasonable description of the fusion processes .
the most successfull calculations performed so far are based on the cdcc method .
the major lesson we have learnt from them is that continuum - continuum couplings play an essential role in the fusion processes , reducing substantially the cf cross section .
this result arises from the statistical irreversibility of transitions to the continuum .
we have pointed out that the presently available cdcc calculations have the shortcoming of not providing separate descriptions of the various fusion processes following breakup .
this is a very hard task , which might be handled in a better way in classical and semiclassical models . | we review the theoretical methods to evaluate fusion cross sections in collisions of weakly bound nuclei .
we point out that in such collisions the coupling to the breakup channel leads to the appearance of different fusion processes .
the extension of the coupled - channel method to coupling with the continuum is the most successful treatment for these collisions .
however , evaluating separate cross section for each fusion process remains a very hard task . |
the investigation of coherent radio emission from the pulsar magnetosphere has attracted a great deal of attention ( cordes 1979 ; michel 1982 ; asso et al . 1990 ) .
a powerful collective emission occurs when relativistic electron beams with density @xmath0 1 per cent of the pair plasma density , scatter off coherently from concentrations of plasma waves ( cavitons ) ( benford 1992 ) .
the role of collective plasma processes in the absorption and spectral modification of the radio waves is well known ( e.g. , beal 1990 ; krishan & wiita 1990 ; benford 1992 ; gangadhara & krishan 1992 , 1993 , 1995 ; gangadhara , krishan & shukla , 1993 ) .
von hoensbroech , lesch and kunzl ( 1998 ) have demonstrated that degree of linear polarization decreases with increasing frequency while the degree of circular polarization shows the opposite trend . in this paper
, we estimate the role of stimulated raman scattering on the polarization of electromagnetic waves propagating in the pulsar magnetosphere .
we assume that the physical conditions in the pulsar magnetosphere are those of the classical standard model ( ruderman & sutherland 1975 ) which describes the generation of ultrarelatvistic beams of electrons / positrons and the creation of the pair plasma .
the beams and the pair plasma are in relativistic motion along the bundle of open magnetic field lines that delimit the active region of magnetosphere .
the stimulated raman scattering is considered as a parametric decay of the initial transverse electromagnetic ( pump ) wave into another electromagnetic wave and a longitudinal plasma wave .
the physics of stimulated raman scattering in a plasma has been explained in many papers and books ( e.g. , drake et al . 1974 ; liu & kaw 1976 ; hasegawa 1978 ; kruer 1988 ) . in two ways , stimulated raman scattering may be important in the pulsar environment : first , it may act as an effective damping mechanism for the electromagnetic waves generated by some emission mechanism at the lower altitudes in the pulsar magnetosphere . at those altitudes , the resonant conditions for stimulated raman scattering , i.e.
, the frequency and wave number matching might not be satisfied .
this results in a short time variability which is generally observed in pulsar radio emission .
secondly , it may provide an effective saturation mechanism for the growth of the electromagnetic waves provided that the conditions for the wave excitation by some mechanism are satisfied in the region where an effective stimulated raman scattering can take place .
the first case can be simplified by treating the intensity of the pump is constant in time .
then the nonlinear equations , which describe the wave coupling , become linear in amplitudes of the decaying waves , and the exponentially growing solutions will imply an effective energy transfer from the pump wave .
this approximation breaks down when the amplitudes of the decay waves become comparable to the pump wave or when the amplitudes of the decay waves enter nonlinear stage , and start loosing energy due to some nonlinear processes such as wave - particle trapping and acceleration .
the second case is more complicated , where stimulated raman scattering acts as a nonlinear saturation mechanism and the amplitudes of all waves may be of the same order .
this case can be considerably simplified when damping of the plasma wave is very strong or if it leaves the region of the resonant interaction fast enough .
we neglect the nonlinear stages of stimulated raman scattering and the development of langmuir turbulence , which leads to wave particle trapping or quasilinear diffusion .
if the pump is monochromatic , the growth rate of stimulated raman scattering can become very high , as in conventional laboratory laser - plasma interaction .
however , in the case of pulsars the pump can be broadband , and in the limit where the bandwidth @xmath1 of the pump wave is much larger than the growth rate of stimulated raman scattering , we can use a random phase approximation for the statistical description of the interacting waves .
tsytovich and shvartsburg ( 1966 ) have given a general expression for the third order nonlinear currents excited in a magnetized plasma .
since the corresponding expressions are very complicated , the general case of raman scattering becomes very difficult to consider .
however , one can make some useful simplifications , when considering stimulated raman scattering in the pulsar magnetosphere .
first , in superstrong magnetic field , we can expand the currents in @xmath2 where @xmath3 is the cyclotron frequency .
second , if the pair plasma has the same distributions for electrons and positrons then some of the nonlinear currents cancel out , as they are proportional to the third power of the electric charge ( this cancellation is exact in the unmagnetized electron - positron plasma ) .
third , all the three interacting waves propagate along magnetic field .
this is an important but less justified approximation .
it allows us to simplify the problem considerably , and to obtain a dispersion relation for stimulated raman scattering .
the polarization of the pulsar signals appear to bear critically on the pulsar radio emission process and the emission beam geometry .
one or more reversals of the sense of circular polarization has been observed in the intergrated profiles of several pulsars .
however , in individual pulses circular polarization changes sense many times across the pulse window .
further , it is important to determine whether the depolarization is a geometric effect or results from radiation plasma interactions .
there have only been very preliminary attempts to explain depolarization and microvariability using plasma mechanisms ( benford 1992 ) .
our purpose in this paper is to derive the properties of natural plasma modes and to explore some possible implications concerning the interpretation of the observed polarization , notably large and symmetric circular polarization in some pulsars ( e.g. , psr b0628 - 28 , and 1914 + 13 ) . in 2 , we derive the dispersion relation for stimulated raman scattering of an electromagnetic wave , and give an analytical expression for the growth rate of the instability . in 3 , we define the stokes parameters and compute the polarization states of the scattered electromagnetic wave .
the discussion of our findings is given in 4 .
we begin with a model consisting of a pulsar with nonthermal component of radiation interacting with plasma in the emission region at a distance @xmath4 cm , ( neutron star radius @xmath5 km ) from the neutron star surface , where the magnetic field is about @xmath6 g. plasma particles may be all in their lowest landau level with no larmor gyration , however , plasma can have one - dimensional distributions of momenta along the magnetic field ( blandford 1975 ; lominadze , machabelli & usov 1983 ) .
this is because the synchrotron loss time for the decay of the perpendicular component of momentum is always short compared to the transit time at the stellar surface for any velocity .
the simplest model for the plasma is density declining in proportion to @xmath7 with no gradients in the distribution functions across @xmath8 the nonlinear interaction of radiation with plasma causes the modulation instability leading to enhancement of nonresonant density perturbations and the radiation amplification by free - electron - maser , which produces intense electromagnetic waves ( lesch , gil & shukla 1994 ; gangadhara , krishan & shukla 1993 ; gangadhara & krishan 1992 ) .
since the frequency of these electromagnetic waves is close to the plasma frequency , they resonantly couple with the subpulse - associated plasma column in the pulsar magnetosphere and drive the stimulated raman scattering .
consider a large amplitude electromagnetic wave @xmath9 with an electric field : e_i=_xi(k_ir - _it ) _ x+_yi(k_ir - _it+_i)_y interacts with the plasma in the pulsar magnetosphere .
we follow ruderman & sutherland s ( 1975 ) approach to estimate the density and plasma frequency of the plasma moving within the bundle of field lines . for the typical parameters : lorentz factor of primary particles @xmath10 and for pair plasma particles
@xmath11 , magnetic field @xmath12 g , and pulsar period @xmath13 s , we obtain the particle number density n_o= = ^-3 and the plasma frequency _p=2_()^1/2 = ^-1
. the plasma in the emission region of pulsar magnetosphere may be birefringent ( melrose & stoneham 1977 ; melrose 1979 ; barnard & arons 1986 ; von hoensbroech , lesch and kunzl 1998 ) . in these models ,
two modes of wave propagation is generally allowed in magnetoactive plasma : one mode is polarized in the @xmath14@xmath15 plane and other mode in the direction perpendicular to it . following these models , we resolve @xmath16 into two modes @xmath17 and @xmath18 such that they are polarized in the directions parallel and perpendicular to the @xmath14@xmath15 plane , respectively .
it is well known from laser - plasma interactions that large amplitude electromagnetic waves resonantly interact with the plasma when the radiation frequency becomes close to the plasma frequency . since the two modes have different indexes of refraction ( mckinnon 1997 ) , i.e. , the response of plasma is not same for the two radiation modes , it is reasonable to assume that @xmath19 couples with the density perturbation @xmath20 and @xmath21 couples with @xmath22 in the plasma medium . since the ponderomotive force is proportional to @xmath23 and @xmath24 ,
the coupling between the radiation and the density perturbations is nonlinear .
hence the density perturbations grow up which lead to currents and mixed electromagnetic electrostatic side - band modes at ( @xmath25 in turn these side - band modes couple with the incident wave field , producing a much stronger ponderomotive force , which amplifies the original density perturbation . hence a positive feedback system sets in , which leads to an instability . the electric field @xmath26 of the electromagnetic wave scattered through an angle @xmath27 with respect to @xmath14
can be written as e_s=_xs(k_sr - _st)_x+_ys ( k_sr - _st + _ s)_y .
the propagation directions of incident wave @xmath28 and scattered wave @xmath29 are illustrated in fig . 1
, such that @xmath30 @xmath31 and @xmath32 the primed coordinate system is rotated through an angle @xmath27 about the y - axis
. then the scattered wave in the unprimed coordinate system is given by e_s=_xs ( _ s ) ( k_sr - _st ) _ x + _ ys ( k_sr - _st + _
s)_y -_zs(_s ) ( k_sr - _st)_z , where @xmath33 the quiver velocity @xmath34 of positrons and electrons due to the radiation fields @xmath16 and @xmath26 is given by = ( e_i+e_s ) , where @xmath35 and @xmath36 are the charge and rest mass of the plasma particle .
-2.5 cm 2.5 cm 15.0 truecm -7 cm the wave equation for the scattered electromagnetic wave is given by ( ^2 - 1 ^ 2)e_s=4j , where @xmath37 is the velocity of light and @xmath38 is the current density .
the components of equation ( 7 ) are d_s_xs(_s)(k_s r - _s t ) = - _ xin_1 ( k_s r - _s t ) and d_s _ ys ( k_s r - _s t+_s ) = - _ yin_2 ( k_s r - _s t+_i- ) and d_s _ zs(_s ) ( k_s r - _s t ) = 0 , where @xmath39 and _s = _i - , k_s = k_i - k . in quantum language these two relations may be interpreted as the conservation of energy and the momentum along the magnetic field , respectively .
when these conditions are satisfied , stimulated raman scattering is exicted resonantly and the expression @xmath40 becomes the dispersion relation of the scattered electromagnetic mode . if we cancel the instantaneous space and time dependent cosine functions on both sides of equation ( 8) , we get d_s_xs_s = - _ xin_1 .
similar to the phase matching conditions ( equation 11 ) , equation ( 9 ) gives the condition between the initial phases : _
s=_i- , and hence we obtain d_s_ys=-_yin_2 . dividing equation ( 14 ) by equation ( 12 ) , we have _ s=_i ( _ s ) , where @xmath41 @xmath42 and @xmath43 the value of @xmath44 is determined by @xmath16 in such way that @xmath45 couples with @xmath19 and @xmath46 couples with @xmath21 .
we consider the vlasov equation to find the low frequency plasma response : + v.f+1m_0 ( e-).f=0 , where @xmath47 is the scalar potential associated with the electrostatic waves , @xmath48 is the particle distribution function and @xmath49 is the ponderomotive potential
. using @xmath50 we can linearize equation ( 16 ) , and obtain : + ( f_2)+ v.(f_1 ) + v.(f_2)+1 ( e-).f_0=0 , where @xmath51 and @xmath52 the ponderomotive force of the radiation field depends quadratically on the amplitude of the electric field . physically , it is a radiation pressure which amplifies the density perturbations by exciting the slow longitudinal fields and motions .
the ponderomotive potential is given by : @xmath53\eps_i\eps_s,\end{aligned}\ ] ] where the bracket @xmath54 represents the average over the fast time scale @xmath55 in the presence of a strong magnetic field , the pair plasma becomes polarized and one dimensional .
if there is some relativistic drift between the electrons and the positrons then the ponderomotive force becomes effective , which will induce the nonlinear density perturbations ( asso 1993 ) : n_- , where @xmath56 is the envelope of the parallel electric field of langmuir wave which is slowly varying with space and time , and @xmath57 is the plasma temperature . using these density perturbations and poisson equation , we can self
consistently derive the potential @xmath58 : = -(n_1+n_2 ) .
now , substituting the expressions for @xmath59 @xmath60 @xmath61 @xmath62 @xmath58 and @xmath63 into equation ( 17 ) and using the condition ( 13 ) , we obtain f_2 + f_1+=0 , where @xmath64 for @xmath65 and @xmath66 @xmath67 while at the other valuse of @xmath68 we have to find the average @xmath69 over the time scale @xmath70 therefore , we have f_2+f_1 _ a=- , where @xmath71 the sum of density perturbations @xmath72 can be estimated as n_2+n_1_a = _ -^ n_0 ( f_2+f_1_a ) dv = - , where = _
-^ k.f_0dv , is the susceptibility function ( liu & kaw 1976 ; fried & conte 1961 ) .
using @xmath73 we can write ( 1 + ) ( + _ a ) n_1 = - a. multiplying equation ( 12 ) by @xmath74 and equation ( 14 ) by @xmath75 , and adding we get ( _ s_i+_a_s)_s=- _ i(_a+_i^2 ) , where @xmath76 and @xmath77 now , using equations ( 25 ) and ( 26 ) , we obtain the dispersion relation for stimulated raman scattering : ( 1 + ) ( + _ a ) = , where @xmath78 is the quiver velocity of plasma particles due to the electric field of incident electromagnetic wave . if @xmath79 erg s@xmath80 is the luminosity of pump radiation with frequency @xmath81 rad s@xmath80 at a distance @xmath82 cm from the source then v_0 = ( ) ^1/2 = 4.310 ^ 9 ^ -1 .
when the phase matching conditions are met , the instability becomes more efficient , and @xmath83 represent the dispersion relation of the stokes mode .
the thermal speed @xmath84 of the plasma can be expressed in terms of its energy spread in the laboratory frame .
let @xmath85 be the velocity of electron - positron plasma , then the velocity spread @xmath86 is given by ( hasegawa 1978 ; gangadhara & krishan 1992 ) v_c 1 .
now , using the lorentz transformation of velocity , we can show that v_=v_z= = because @xmath87 hence the thermal speed , in the plasma frame , is given by v_t = c .
for @xmath88 we get @xmath89 cm s@xmath90 if we separate equation ( 27 ) into real and imaginary parts , we get two coupled simultaneous equations . by solving them numerically , we find the growth rate @xmath91 of stimulated raman scattering .
figure 2 shows the growth rate as function of @xmath92 and @xmath93 since the plasma density decreases with r as @xmath94 growth rate decreases as @xmath92 increases . also
, if the radiation frequency becomes higher than the plasma frequency , then radiation and plasma do not couple resonantly , which leads to the decrease of growth rate at higher values of @xmath93 30 cm .
truecm -2 cm * figure 2 . * the growth rate @xmath91 of stimulated raman scattering is plotted with respect to @xmath92 and @xmath93 0.5 cm to study the variation of @xmath91 in the @xmath95@xmath96 plane , we have made a contour plot ( fig .
3 ) , the labels on the contours indicate the values of log@xmath97 debye length increases with the increase of @xmath98 therefore , plasma wave experiences a large landau damping , which leads to the drop in growth rate . the parameter @xmath95 is the ratio of amplitudes of electromagnetic waves , which are polarized in the directions parallel and perpendicular to the @xmath99 plane .
when @xmath95 is small , the wave polarized in the direction parallel to @xmath99 plane becomes strong and will have a component along @xmath8 therefore , the coupling between radiation and plasma will be strong , which leads to the higher growth rate at the smaller values of @xmath100 11.0 truecm -1.5 cm * figure 3 .
* contour plot of @xmath91 in the @xmath96@xmath95 plane .
the labels on the contours indicate the log(@xmath91 ) values .
0.5 cm for @xmath101 stokes mode becomes resonant , and anti - stokes mode becomes non resonant .
then the equation ( 27 ) can be written as ( - _l+i_l)( - _l+i_s)(+_a)=-10 ^ 18 s^-2 , where @xmath102 and _ l=+_c is the damping rate of the plasma wave , and the debye length @xmath103 for @xmath104 we find @xmath105 s@xmath90 the pair plasma collision frequency @xmath106 s@xmath80 and coulomb logarithm @xmath107 the collisional damping rate of the scattered electromagnetic wave is given by @xmath108 s@xmath90 for @xmath109 plasma waves are highly damped and hence plasma losses its collective behaviour .
therefore , stimulated raman scattering changes into the induced compton scattering of electromagnetic waves by the plasma particles .
in the conventional treatment of induced compton scattering in pulsars ( e.g. , blandford & scharlemann 1976 ; sincell & krolik 1992 ) the collective effects of the plasma are ignored .
the collective treatment of the wave scattering by plasma particles is justified if the condition @xmath110 or @xmath111 is met , where @xmath112 is the phase velocity of the plasma wave .
this condition implies that the wave number of oscillation of the electrons in the beat wave of the incident and scattered waves be much less than the inverse of the debye length .
it is not satisfied , when the beat wave is strongly landu damped ( @xmath113 ) or the beat wave will not feel the presence of a medium ( @xmath114 ) , so that the scattering process will be described as induced compton scattering .
when @xmath115 by setting @xmath116 we can solve equation ( 32 ) for the growth rate : = -12(_l+_s ) . stimulated
raman scattering is a threshold process : if the intensity of the pump exceeds the threshold , then only it would start converting its energy into the decay waves .
the threshold condition for the excitation of stimulated raman scattering is given by ( ) _ thr= 9.3 10 ^ -19 _
10r_8 ^ 3/2 _ l_s .
the typical threshold intensities for stimulated raman scattering are of the order of the observed intensities , implying that pulsar magnetosphere may be optically thick to raman scattering of electromagnetic waves .
the growth rate just above the threshold is given by = 4.310 ^ 8 ^-1 , which is proportional to @xmath117 the maximum growth rate attainable for @xmath118 on the other hand , is = 10 ^ 9 ^ -1 .
when the phase matching conditions are satisfied , the growth rate @xmath119 becomes large , and the scattered mode is amplified and become a normal mode of the plasma . under some conditions ,
the scattered mode leaves the plasma with polarization which may be different from the polarization of the incident wave .
the polarization states of the incident and scattered waves can be described more accurately using the stokes parameters ( rybicki & lightman 1979 ) : i_j= _ xj_xj^*+_yj_yj^ * , q_j= _ xj_xj^*-_yj_yj^ * , u_j= 2_xj_yj^*(_j ) and v_j= 2_xj_yj^*(_j ) , where @xmath120 for the incident wave , and @xmath121 for the scattered wave .
the linear polarization is given by l= , and the polarization position angle is given by _ j=(u_j
. the transfer of energy between the modes will be efficient only when the energy of the pump wave is strong enough to overcome the damping losses or escape of the generated waves . using the manley
rowe relation ( weiland & wilhelmsson 1977 ) = i_s_s , we find the relation between the incident flux @xmath122 and the scattered flux @xmath123 : i_s=(1-_i)i_i . in the following two subsections , we consider the cases where the polarization state of the incident wave is linear and circular , and compute the polarization states of the scattered mode . consider a linearly polarized electromagnetic wave ( @xmath124 @xmath125 and @xmath126 which excites stimulated raman scattering in the magnetospheric plasma . using the plasma and magnetic field parameters discussed in the previous section , we computed the growth rate of stimulated raman scattering . for @xmath127
the instability becomes quite strong and the landau damping of langmuir wave is minimal .
hence , the stimulated raman scattering is resonantly excited .
figure 4 shows the behaviour of linear ( solid line ) and circular ( broken line ) polarization of scattered mode with respect to @xmath44 .
it shows at some values of @xmath128 which are close to 0.1 , the linear polarization of the incident wave can be converted almost completely into circular polarization of scattered wave .
charge density variations within the pulse window could enhance the conversion efficiency of linear to circular polarization to vary .
-1.0 cm 10 .
truecm -0.5 cm * figure 4 . * the solid line and
broken line curves indicate the variation of linear ( @xmath129 ) and circular ( @xmath130 ) polarization of the scattered mode with respect to @xmath131 the normalizing parameter @xmath123 is the intensity of scattered mode .
0.5 cm the variation of polarization angle of the scattered wave with respect to @xmath128 at different values of @xmath132 is shown in fig . 5 .
for @xmath133 and @xmath134 the scattered mode becomes orthogonally polarized with respect to the incident wave .
furthermore , if there is any variation in the plasma density or radiation intensity , the value of @xmath44 fluctuates and the scattered modes produced with @xmath135 become orthogonally polarized with respect to those produced with @xmath136 -1.5 cm 10.5 truecm -1 .
cm * figure 5 .
* polarization angle @xmath137 of the scattered mode is plotted as a function of @xmath128 at different values of @xmath138 @xmath139 & @xmath140 suppose that the incident wave is circularly polarized ( @xmath141 and @xmath142 then the scattered mode will be linearly polarized for @xmath143 while at the other values , both linear and circular polarizations with different proportion exist , as indicated by fig .
6 . for @xmath144 and @xmath145 the sense of circular polarization of the scattered mode becomes opposite to that of the incident wave . hence depending upon the plasma and radiation conditions , stimulated raman scattering can change the polarization of the pulsar radio signals .
cm 10.5 truecm -0.5 cm * figure 6 . * the variation of linear ( @xmath129 ) and circular ( @xmath130 ) polarization of the scattered mode , with respect to @xmath128 are indicated by the solid line and broken line curves .
the normalizing parameter @xmath123 is the intensity of the scattered mode .
the variable nature of circular polarization is evident from the polarization distributions by manchester , taylor and huguenin ( 1975 ) , backer and rankin ( 1980 ) and stinebring et al .
( 1984a , b ) .
very high degrees of circular polarization are occasionally observed in individual pulses , even up to 100 per cent ( cognard et al . 1996 ) , while the integrated or average pulse profiles generally indicate much smaller degree of circular polarization ( e.g. , lyne , smith & graham 1971 ) , which show that the sign of circular polarization is variable at any given pulse phase .
radhakrishnan and rankin ( 1990 ) have identified two extreme types of circular polarization in the observations : ( a ) an antisymmetric type wherein the circular polarization changes sense in midpulse , and ( b ) a symmetric type wherein it is predominantly of one sense .
the correlation of sense of antisymmetric type of circular polarization with the polarization angle swing indicates the geometric property of emission process , and is highly suggestive of curvature radiation
. the diverse nature of circular polarization may be the consequence of pulsar emission mechanism and the subsequent propagation effects in the pulsar magnetosphere ( e.g. , melrose 1995 ; von hoensbroech , lesch and kunzl 1998 ) .
it seems rather difficult to explain the various circular polarization behaviours using the widely accepted magnetic pole models ( e.g. , radhakrishnan & cooke 1969 ; komesaroff 1970 ; sturrock 1971 ; ruderman & sutherland 1975 ) . the symmetric type of circular polarization observed in some pulsars ( e.g. , psr b1914 + 13 , 0628 - 28 ) as shown in figs . 7 and 8 , may be due to the propagation effects .
however , it seems difficult for propagations effects to explain , how the sign of the circular polarization can change precisely at the center of the pulse in the case of antisymmetric type , as seen in many pulsars e.g. , psr b1859 + 03 and b1933 + 16 ( rankin , stinebring & weisberg 1989 ) . 0.5 cm 10.0 truecm 0.3 cm * figure 7 . * psr b0628 - 28 , an example of a pulsar with ` symmetric ' circular polarization and high linear polarization ( from lyne & manchester 1988 ) .
0.5 cm if we are to understand the radio emission mechanism , we must understand the physical state of the radio - loud plasma in the polar cap .
it is this plasma that is the site of instabilities which are thought to produce coherent radio emission .
the role of different propagation effects on the pulsar polarization has been discussed by cheng and ruderman ( 1979 ) , beskin , gurevich and istomin ( 1988 ) and istomin ( 1992 ) .
the mechanisms proposed by these authors predict a frequency dependence for circular polarization , with weaker polarization at higher frequencies .
this is seen in some pulsars ( e.g. , psr b0835 - 41 , 1749 - 28 , 1240 - 64 ) but it is not generally the case ( han et al .
istomin suggested that the linearly polarized incident wave becomes circularly polarized as a result of generalized faraday rotation , however , it is observationally known that no generalized faraday rotation is evident in pulsar magnetospheres ( cordes 1983 ; lyne & smith 1990 ) .
. -1.0 truecm 11.3 truecm 0.5 cm * figure 8 .
* polarization of psr b1914 + 13 , a pulsar with strong circular polarization over the whole observed profile ( from rankin , stinebring & weisberg 1989 ) .
0.5 cm we have presented a model to explain the polarization changes due to the propagation of radio waves through the magnetospheric plasma . the features like a large change in polarization angle , sense reversal of circular polarization and extremely rapid temporal changes in intensity would help us to explain many observations , for which , the existing mechanisms proved to be inadequate . because of the very strong dependence of polarization angle on plasma parameters via the growth rate , in an inhomogeneous plasma medium the depolarization is a natural outcome .
we believe that the plasma process such as the stimulated raman scattering may be a potential mechanism for the polarization variability in pulsars .
the circular polarization of a number of pulsars varies with frequency .
the two clear examples are psr b1240 - 64 and 2048 - 78 , from which the opposite senses or transitions of circular polarization have been observed at lower and higher frequencies ( han et al .
we considered the stimulated raman scattering of the transverse electromagnetic waves in the electron - positron plasma of pulsar magnetosphere .
the value of radiation plasma coupling parameter @xmath44 is determined by the polarization of the incident wave , and its value can be determined only by the nonlinear analysis . in the process of three
wave interaction , the phase matching condition ( equation 13 ) between the initial phases @xmath146 @xmath147 @xmath148 and the value of @xmath44 determine the polarization state of the scattered wave .
many short time scale polarization variabilities , in individual pulses , such as variations in amount of linear and circular polarization , sense reversal of circular polarization and polarization angle swings can be accounted for by considering the stimulated raman scattering .
the time scales over which the changes takes place , in individual pulses , are of the order of e - folding time , which is the inverse of the growth rate of stimulated raman scattering .
it seems rather unlikely that the diverse behaviour of circular polarization can be accounted for by a single mechanism .
both intrinsic emission and propagation effects seem possible .
the strong symmetric type of circular polarization observed in some pulsars is most probably generated by propagation effects , such as the stimulated raman scattering .
further simultaneous observations over a wide frequency range would be valuable in sorting out the importance of propagation effects .
we are grateful to y. gupta and a. von hoensbroech for discussions and comments .
0.2 cm asso e. 1993 , mnras 264 , 940 asso e. , pelleiter g. , sol h. 1990 , mnras 247 , 529 backer d.c . , rankin j.m . 1980 ,
apjs 42 , 143 barnard j.j . , arons j. 1986 , apj 302 , 138 beal j.h . 1990 , in physical processes in hot cosmic plasmas , eds .
brinkmann , w. , et al .
, ( kluwer academic publications ) , 341 benford g. 1992 , apj 391 , l59 beskin v.s . ,
gurevich a.v .
, istomin ya.n .
1988 , ap .
space sci .
146 , 205 blandford r.d .
1975 , mnras 170 , 551 blandford r.d .
, scharlemann e.t .
1976 , mnras 174 , 59 cheng a.f . ,
ruderman m.a .
1979 , apj 229 , 348 cognard i. , shrauner j.a . , taylor j.h .
, thorsett s.e .
1996 , apj 457 , l81 cordes j.m .
1979 , space sci .
24 , 567 cordes j.m .
1983 , in positron
electron pairs in astrophysics , ed .
burns m.l.,harding a.k . , ramaty r. , aip conf . proc . 101 , 98 drake j.f . , et al .
1974 , phys .
fluids 17 , 778 fried d. , conte s.d .
1961 , the plasma dispersion function , ( academic press , new york ) , 1 gangadhara r.t .
, krishan v. 1992 , mnras 256 , 111 gangadhara r.t . ,
krishan v. 1993 , apj 415 , 505 gangadhara r.t .
, krishan v. 1995 , apj 440 , 116 gangadhara r.t . , krishan v. , shukla p.k . 1993 ,
mnras 262 , 151 han j.l . , manchester r.n .
, xu r.x . , qiao g.j . 1998 ,
mnras , ( submitted ) .
von hoensbroech a. , lesch h. , kunzl t. , 1998 , a&a 336 , 209 hasegawa a. 1978 , bell system tech . j. 57 , 3069 istomin ya.n .
1992 , in hankins t.h . ,
rankin j.m .
, gil j.a .
, eds , the magnetosphere structure and emission mechanism of radio pulsars , proc .
128 , 375 komesaroff m.m .
1970 , nat .
225 , 612 krishan v. , wiita p.j . 1990 , mnras 246 , 597 kruer w.l .
1988 , the physics of laser - plasma interactions , ( addison - wesley , new york ) , 70 lesch h. , gil j.a . , shukla p.k .
1994 , space sci .
68 , 349 liu c.s .
, kaw p.k .
1976 , advances in plasma phys . ,
simon , a. , & thompson , w. b. , ( interscience new york ) , 6 , 83 lominadze dzh.g .
, machabeli g.z . , usov v.v .
1983 , apss 90 , 19 lyne a.g . ,
manchester r.n . , 1988 ,
mnras 234 , 477 lyne a.g . ,
smith f.g .
1990 , pulsar astronomy , cambridge univ .
press , cambridge , 234 lyne a.g . , smith f.g . ,
graham d.a .
1971 , mnras 153 , 337 manchester r.n .
, taylor j.h . , huguenin g.r .
1975 , apj 196 , 83 mckinnon 1997 , apj 475 , 763 melrose d.b .
1979 , aust .
32 , 61 melrose d.b .
1995 , jaa 16 , 137 melrose d.b .
, stoneham 1977 , proc .
astron . soc .
australia 3 , 120 michel f.c .
1982 , rev .
54 , 1 radhakrishnan v. , cooke d.j . 1969 , astrophys .
lett . 3 , 225 tsytovich v.n . , shvartsburg a.b .
1966 , sov .
jetp 22 , 554 radhakrishnan v. , rankin j.m . 1990 , apj 352 , 258 rankin j.m . , campbell d.b . ,
backer d.c .
1974 , apj 188 , 608 rankin j.m . , stinebring d.r . ,
weisberg j.m .
1989 , apj 346 , 869 ruderman m. , sutherland p. 1975
, apj 196 , 51 rybicki g.b .
, lightman a.p .
1979 , radiative processes in astrophysics , ( a wiley interscience publication ) , 62 sincell m.w . , krolik j.h .
1992 , apj 395 , 553 stinebring d.r . ,
et al . 1984a , apjs 55 , 247 stinebring d.r .
, et al . 1984b , apjs 55 , 279 sturrock p.a .
1971 , apj 164 , 529 weiland j. , wilhelmsson h. 1977 , coherent non linear interaction of waves in plasmas , ( pergamon press ) , 60 | we consider the role of a propagation effect such as the stimulated raman scattering on the polarization of radio pulses .
when an intense electromagnetic wave with frequency close to the plasma frequency interacts with the plasma in the pulsar magnetosphere , the incident wave undergoes stimulated raman scattering . using typical plasma and magnetic field parameters ,
we compute the growth rate and estimate the polarization properties of the scattered mode . at some conditions , we find that the polarization properties of the scattered mode can become significantly different from those of the incident wave . the frequencies , at which strong raman scattering occurs in the outer parts of magnetosphere , fall into the observed radio band . in some pulsars , for example , psr b0628 - 28 and 1914 + 13 , a large and symmetric type of circular polarization has been observed .
we propose that such an unusual circular polarization is produced by the propagation effects . -2
cm mpp.sty epsf.sty stars : pulsars - plasmas - waves - radiative transfer |
the following paper is dedicated to nihat berker and describes a mathematical curiosity that i stumbled across a few years ago .
it brings together two ideas that were a daily topic of discussion with nihat during my graduate student days : duality and renormalization .
the work also resulted from my admiration of the elegant recursive algorithms that lie at the basis of the efficient solution of a number of practical problems , most notably sorting and the fourier transform . for a transitionally invariant problem ,
a fourier transform usually either solves or at least decouples the spatial degrees of freedom so that the original problem becomes simple .
when applied directly on data of length @xmath0 it uses a number of operations that scale like @xmath1 . the `` fast fourier transform '' ( fft )
dramatically reduces the number of arithmetic operations using a clever recursive algorithm . an interesting question would be if if a single fft recursion could somehow be the basis of a renormalization transformation that would partially decouple a problem at each stage of recursion .
ultimately i was not able to answer this question but in the process came up with a surprising result ; that repeated application of the fft recursion can map a problem from one to two dimensions .
before proceeding to the calculation this i make a brief comment on the history of the fft . in 1965
cooley and tukey wrote a famous paper entitled _ `` an algorithm for the machine calculation of complex fourier series '' _ .
cooley and tukey cited i.j .
good as having influenced their work .
the impact of the `` cooley - tukey algorithm '' was enormous and their `` fast fourier transform '' immediately became indispensable tools for engineers and scientists as computers started to become widely available at this time .
these days the `` fft '' algorithm is part of every numerical package and most of us do nt pay much attention to the algorithm . soon after the paper came out it
was realized that the `` fft '' algorithm had been repeatedly rediscovered earlier . a detailed historical discussion on the subject is available by heideman , johnson and burrus .
the `` fft '' was eventually traced back to c.f .
it appears certain that he discovered the algorithm around 1807 but did not publish it .
given the impossibility of making significant numerical calculations at the time gauss probably did not consider the result particularly important .
the algorithm was eventually published posthumously in his collected works .
thus gauss original work was done approximately simultaneously with fourier s development of his famous trigonometric expansion . in my opinion
though , cooley and tukey deserve great credit for realizing in 1965 the importance of an efficient numerical fourier transform and deriving the algorithm at a time when it became useful .
the power of the fft algorithm is that it dramatically reduces the number of arithmetic operations required for a fourier transform of a large data set . for the most common `` radix-2 '' algorithm
the number of data points is taken to be @xmath2 .
the number of operations will be on the order of @xmath3 rather than the @xmath4 number of operations that results from applying a simple matrix operation .
the fft algorithm can also be implemented on arbitrary length data where the number of arithmetic operations scales like @xmath5 times the sum of prime factors of @xmath5 .
when @xmath5 is prime , the fft reduces to the ordinary matrix operation . without going into detail , i will now write down the key formulas using radix-2 algorithm that shows how the fft is based on recursion .
we begin with data @xmath6 where @xmath7 and @xmath8 .
we define the fourier transform @xmath9 the sum can be split into two parts : @xmath10 where @xmath11 can be seen to be the fourier transform of the even indexed subset of @xmath12 and @xmath13 is the fourier transform of the odd subset .
this procedure can therefore be done recursively , splitting @xmath11 and @xmath13 successively into into odd and even subsets each with half as many data points .
this notation is clumsy and does not lend itself to recursion . to implement it
systematically we must keep track of which order of even and odd blocks we use to construct @xmath12 as we apply the algorithm .
we use the label @xmath14 keep track of our recursive indexing .
we begin by defining @xmath15 to consist of the original data .
the recursion then consists of breaking the first index @xmath16 into odd and even blocks .
since i will shortly write down a precise formula , i will first just the sketch the idea of constructing the indexing .
we lay the `` odd '' and `` even '' blocks next to one another and put in a binary index `` 0 '' last if the block is even and `` 1 '' if the block is odd .
the original block is periodic in length @xmath5 ; each new block is periodic in length @xmath17 and the number of blocks , say @xmath0 in number is doubled to length @xmath18 .
thus the original index @xmath19 will be replaced by an index @xmath20 where the `` block '' index @xmath21 encodes odd and even block and @xmath22 describes the new order in the subblock . doing this recursively
, the index @xmath21 will gradually become the `` frequency '' or `` momentum '' index .
the recursion is finished when the entire space index has been reduced to length one and the momentum index has become length @xmath5 .
let us use the notation @xmath23 to denote the integer @xmath24 modulo @xmath5 .
we encode the algorithm as follows : @xmath25 the inverse is given by @xmath26 it is straightforward to apply this operation to quantum mechanical operators @xmath27 where @xmath28 is simply the one particle ( destruction ) operator on an @xmath29 lattice of `` pseudo real space '' dimension @xmath5 and `` pseudo k - space '' dimension @xmath30 . for the initial iteration
@xmath5 is the original space dimension , and @xmath31 .
equivalently we can simply interpret @xmath32 as the column vector being transformed and @xmath33 as a row vector .
thus we begin with a periodic `` real space '' lattice of length @xmath34 to which we initially attach a bogus `` momentum space '' lattice of length @xmath35 to begin the recursion with @xmath36 . at each fft
step we halve the length of the real space and double the length of the momentum space .
eventually when we are done applying the recursion @xmath37 times the `` real space '' lattice will have shrunk to length one and the variables @xmath38 that correspond to the momentum space lattice will be the fourier transform of the original variables .
what if we stop this recursion `` half way '' ?
provided @xmath5 is @xmath39 where @xmath40 is integer , we will obtain a two dimensional periodic @xmath41 lattice index when we stop the procedure after @xmath40 recursions and thus will have obtained a mapping to a `` two - dimensional '' @xmath41 index @xmath14 on a torus from our original @xmath39 periodic index .
clearly this transformation is highly nonlocal since it mixes fourier components of indexes very far apart . if we start with an arbitrary real space operator written in terms of the real space operators @xmath42 and apply this prescription it is tempting to conclude that we will always get a terrible mess during the recursion .
this is indeed true generically .
however , we know that a transitionally invariant operator is diagonalized by the fourier transform , i.e. written as a simple local operator in the fully transformed basis
. thus eventually after recursion the momentum space blocks must decouple and the intuition that the problem gets more and more intractable under recursion is violated .
furthermore , a model which is invariant under fourier transform should be symmetric in momentum and space coordinates half way through the recursion , so perhaps such a model will retain a structure during the intermediate steps of the recursion as well .
it is precisely question we will explore further by example . in the rest of this communication , i will focus on the following model under the `` partial '' fft operation .
we begin with @xmath43 this is in fact a special case of `` harper s model '' where periodicity is exactly one relative to the length of the lattice . a point which i will discuss at the end of this paper .
we know that the first term transforms to the cosine under fourier transform and vice versa . for this choice of parameters
the model is `` self dual '' i.e. transforms to itself under fourier transform .
it can therefore be expected to be symmetric in some way when partially transformed @xmath37 times , where @xmath44 . for illustration
, we begin with a @xmath45 lattice and apply the fft step three times to transform this model numerically to an @xmath46 lattice .
the result of applying this transformation is given by the following hamiltonian @xmath47 where @xmath48 are only nearest neighbor lattice vectors , where @xmath49 and the phases of @xmath50 correspond to a single flux evenly spread throughout the entire lattice .
the value of @xmath51 where @xmath52 is shown in fig .
[ fig : hcentered ] . where @xmath51 are shown . , width=491 ]
the vertical bonds with phase zero represent the mapping of the cosine potential , whereas the horizontal bonds that carry the flux is the transformation of the hopping .
let us look more carefully at the distribution of fluxes . with the single exception of a plaquette in the top row
, we note that the sum of integers around each plaquette is one , so that the product of phases around each plaquette is exactly @xmath53 which i interpret as a single flux quantum evenly spread on the torus .
periodic boundary conditions are fulfilled and the left and right boundary are mapped to each other , likewise for top and bottom . the rule that the sum of phases around each individual plaquette is the sum of phases around the entire boundary
is obtained form the fact that a plaquette on the top row has flux @xmath54 .
an equivalent phase choice of @xmath55 instead of zero for the top link is equivalent to assigning a flux of @xmath56 through each plaquette .
although i do not present further results here , i assert that similarly applying the fft recursion a different number of times result in rectangular lattices all with a total @xmath55 flux threading each periodic lattice . it is interesting to compare the eigenstates of the tight binding model to the corresponding states mapped by the partial fft .
the absolute square of the six lowest eigenstates @xmath57 of of @xmath58 is plotted in fig .
[ fig : onedplots ] . for the six lowest energy eigenstates of @xmath59 .
, width=491 ] these corresponding eigenstates of @xmath60 is plotted in fig .
[ fig : vortexplots ] and thus represents the `` 2-d '' mapping of the `` 1-d '' states in fig .
[ fig : onedplots ] . for the six lowest energy eigenstates of @xmath61 mapped to the two dimensional lattice , width=491 ] the calculations
are displayed here for values of @xmath0 which are not large ; analogous results for larger values of @xmath0 have been computed .
it is well known that an infinite two - dimensional tight binding model in a constant magnetic field maps to a one tight binding model with a periodic or quasiperiodic potential .
in addition to p.g .
harper s paper mentioned previously , there are several other early classic papers on listed in the references . ] in 1976 , r. hofstadter s performed a numerical analysis of the problem and pointed out the beautiful self - similar structure of the energy spectrum .
a number of paper followed in the 80 s that explored this using renormalization group ideas .
the extent to which these are pertinent to the present problem is addressed in this section . for a tight - binding model in the presence of a magnetic field with flux @xmath62 per plaquette
we can introduce a landau gauge and write the hamiltonian as @xmath63 where @xmath64 and @xmath65 .
noting the invariance of the problem for translations along the x - direction we introduce plane waves along the @xmath66 axis and define @xmath67 resulting in a hamiltonian for @xmath68 of the form in @xmath69 if we demand that this hamiltonian be periodic with period @xmath70 we find that that @xmath71 is the ratio of integers .
the flux per square in the 2-d problem eq .
[ eq : harptwod ] is thus a integral multiple of @xmath72 .
is this the transformation we have been doing in the previous section ? if we begin with a @xmath73 periodic two - dimensional lattice and add a constant magnetic field described in landau gauge , we find that @xmath74 must obey @xmath75 with @xmath40 integer and we obtain an equivalent set of @xmath76 one dimensional problems eq .
[ eq : harponed ] with periodicity @xmath77 indexed by the phase @xmath78 . in a conventional formulation of the harper problem
we therefore can not reduce the flux quantum to sum to a single quantum over the entire @xmath79 lattice .
indeed , the flux configuration in the 2-d landau gauge is quite different from that shown in fig . [
fig : vortexplots ] which describes a single flux quantum spread over the lattice and lacks translational invariance in both @xmath66 and @xmath80 direction .
( note the top row vertical bonds is not uniform along @xmath66 . ) a careful analysis of the gauge choice in fig .
[ fig : vortexplots ] or another clever choice of gauge other than landau might possibly be used to show that the fft remapping of the one - dimensional problem of length @xmath81 is in fact harper s problem , but i have not been able to do this .
more digressions on the harper problem would stray further from the main point of the communication which is the presentation of a decoupling scheme of the tight binding model that is in principle more general than harper s problem .
unfortunately i can find no other notrivial example than the relation between the one and two dimensional harper model , but with flux and periodicity reduced from the ordinary formulation of that problem .
these proceedings has provided an opportunity to publish work which i find interesting but admittedly incomplete .
when performing the recursion on the tight binding model only , each fft recursion decouples the spatial degrees of freedom into two coupled blocks .
my original hope was that these transformation might be the basis of a renormalization transformation ; if an interaction could be found that was similarly decoupled , a novel way to solve an interacting model would be found .
unfortunately this goal has not been fulfilled but perhaps someone might be motivated to look at this problem again with the observations presented here .
i thank nihat berker for being my friend , mentor and collaborator during my first three years as a graduate student at harvard .
his boundless energy and enthusiasm helped make these years very happy and productive .
i also thank him for putting the necessary friendly pressure on me to write up this contribution for these proceedings .
i also thank the referee for making several pertinent and important comments that clarified the relation to harper s equation . | it will be shown how to map a simple one - dimensional tight binding model with a cosine potential in one dimension exactly to a two dimensional tight binding model with periodic boundary conditions with the presence of a single flux quantum spread evenly on the torus .
the mapping is is achieved by a partial sequence of `` fast fourier transform '' ( fft ) steps which if completed would be an exact fourier transform of the original model .
each step of the fft recursively maps a tight binding model into two decoupled sublattices of half the lattice length . |
geometric frustration is a term used to describe magnetic systems where all of the microscopic exchange interactions can not be satisfied simultaneously , and as a consequence this class of materials often exhibits exotic magnetic ground states and complex magnetic behavior .
some common architectures characterized by frustration include pyrochlore@xcite , kagome@xcite , and face - centered cubic magnetic sublattices@xcite . due to the magnetic frustration inherent in these materials
, non - collinear magnetic ground states often form that can lead to multiferroic behavior@xcite . in other cases , conventional long - range magnetic order is replaced by short - range alternatives , including spin glasses@xcite , spin liquids@xcite , and spin ices@xcite . recently ,
the family of materials ar@xmath0o@xmath1 ( a @xmath4 ba , sr ; r @xmath4 rare earth ) have been attracting interest as a new series of frustrated magnets .
these systems consist of two crystallographically - inequivalent rare earth sites with very different , distorted octahedral oxygen environments .
each site forms zigzag chains ( or ladders ) of magnetic r atoms running along the @xmath2-axis .
bulk probes , including magnetic susceptibility and heat capacity , have shown that most members of the family have dominant antiferromagnetic ( afm ) exchange interactions and relatively large frustration indices@xcite .
geometric frustration can arise in this structure type if the next nearest neighbor ( nnn ) @xmath7 intrachain exchange interactions ( ladder legs ) are afm and the nearest neighbor ( nn ) @xmath8 intrachain couplings ( ladder rungs ) are of comparable strength . the phase diagram for both the magnetic @xmath9 @xmath4 1/2@xcite and classical@xcite ising - like ,
afm @xmath8-@xmath7 chain models have been determined theoretically . in both cases , a simple up - down - up - down ( nel ) phase and an up - up - down - down ( uudd ) phase
have been predicted depending on the relative values of @xmath8 and @xmath7 . for @xmath9
@xmath4 1/2 systems in zero field , the two phases are separated by a quantum critical point at @xmath7 @xmath4 @xmath8/2 .
an additional canted phase , with finite magnetization , is found between the nel and uudd states in the zero field phase diagram for classical systems . in both the quantum @xmath9 @xmath4 1/2 and
classical spin limits , the calculations predict a field - induced up - up - down ( uud ) phase over a particular @xmath7/@xmath8 range for an applied magnetic field along the ordered spin direction .
the uud spin structure leads to a 1/3 magnetization plateau . despite the extensive theoretical work on the @xmath8-@xmath7 ising chain model
, there are not many examples that can be used to test its predictions , especially in the classical limit . some members of the ar@xmath0o@xmath1 family , in particular srdy@xmath0o@xmath1 and srho@xmath0o@xmath1 , have recently been proposed as model ising @xmath8-@xmath7 chain systems@xcite .
there are also several other classical ising chain candidates in this family with large total angular momenta @xmath10 associated with the magnetic r atoms .
several recent studies have been performed on srr@xmath0o@xmath1 , and both the nel and uudd ordering predicted by the @xmath8-@xmath7 ising chain model have been observed in these classical magnetic spin systems .
however , there are several aspects of the magnetism not explained by the original theories .
for example , srer@xmath0o@xmath1 exhibits long - range nel order@xcite , characteristic of a small @xmath7 , coexisting with diffuse magnetic scattering@xcite .
monte carlo simulations reveal that the pattern from the latter is best described by dominant @xmath7 afm exchange interactions , and this finding is consistent with quasi-1d uudd ordering .
one way the different scattering patterns can coexist is if the two types of zigzag chains have inequivalent @xmath8 and @xmath7 interactions and therefore host different magnetic ground states .
support for this interpretation comes from rietveld refinement of the neutron powder diffraction ( npd ) data , as the best refinements arise when the long - range order is only associated with one type of er zigzag chain@xcite .
a wide range of magnetic behavior is found across the srr@xmath0o@xmath1 series .
srho@xmath0o@xmath1 shows very similar magnetism to the er analogue with coexisting nel and uudd ordered states , although the correlation length of these states along different crystallographic directions is still under debate@xcite . as in the er case ,
the two different structures can likely be associated with the two types of ho zigzag chains .
in contrast , npd shows that sryb@xmath0o@xmath1 is characterized by long - range nel order on both yb sites , although the ordered yb moment is drastically suppressed on one of the two sites@xcite .
finally , srdy@xmath0o@xmath1 shows evidence for quasi-1d , uudd magnetic correlations in zero field down to 50 mk@xcite , while srtm@xmath0o@xmath1 shows no signs of magnetic ordering down to 65 mk@xcite .
the low ( @xmath6 @xmath11 1 k ) or non - existent magnetic ordering temperatures , coupled with the variety of observed behavior , suggest that some combination of competing exchange interactions , single ion anisotropy , magnetic dipole interactions , and low dimensionality play important roles in shaping the magnetic phase diagrams of these materials . in this work , we expand the earlier studies of these zigzag chain systems to band@xmath0o@xmath1 ( @xmath10 @xmath4 9/2 ) through magnetization , heat capacity , and neutron diffraction measurements on polycrystalline samples .
this material has a curie - weiss temperature of -24 k , while the measurements presented in this work indicate a magnetic transition of only @xmath3 @xmath4 1.7 k , indicative of a high magnetic frustration index f @xmath12 12 .
neutron diffraction data show evidence for a long - range antiferromagnetic ground state that arises from only one of the two nd sites , characterized by a propagation vector of @xmath5 @xmath4 ( 0 1/2 1/2 ) and the spins lying in the @xmath13-plane .
furthermore , low temperature magnetization and heat capacity measurements as a function of applied field reveal an intermediate field - induced ordered state , and also indicate that any magnetic order is completely suppressed in an applied field of about 4 t as @xmath14 .
single phase polycrystalline band@xmath0o@xmath1 samples were synthesized by a standard solid - state reaction method from high - purity starting materials of baco@xmath15 and nd@xmath0o@xmath15 .
about 15% extra baco@xmath15 was added to compensate for the evaporation of barium .
the starting materials were mixed and ground in an agate mortar and then pressed into pellets .
the pellets were sintered in a flowing ar atmosphere at 1150@xmath16c for 8 hours .
the resultant pellets were then reground and pressed into new pellets for more sintering in the ar atmosphere at 1300@xmath16c for 10 hours .
the final product was confirmed to be single phase by laboratory x - ray powder diffraction
. neutron powder diffraction ( npd ) was performed with 5 g of polycrystalline band@xmath0o@xmath1 at oak ridge national laboratory using the hb-2a powder diffractometer of the high flux isotope reactor .
the sample was sealed in an al sample can with he exchange gas , and a he-3 sample environment was used for the experiment with a base temperature of 300 mk .
all measurements were conducted with a neutron wavelength of 2.41 and a collimation of 12@xmath17-open-6@xmath17 .
the npd data was analyzed using the rietveld refinement program fullprof@xcite and the representational analysis software sarah@xcite .
the specific heat was measured with a home - built probe based on the adiabatic heat - pulse technique in a 3he/4he dilution refrigerator from oxford instruments .
the experiment was carried out in a 14 t superconducting magnet .
special care was taken to ensure the proper thermalization of the cold - pressed polycrystalline sample during the low temperature measurements .
magnetization measurements in fields up to 7 t were performed in a quantum design mpms squid magnetometer .
data from @xmath12 1.7 k to 100 k was taken using the standard @xmath18he setup .
data between 0.44 k and 2 k was collected with an iquantum @xmath19he insert that fits inside the mpms sample space and utilizes the same squid detection system and procedure .
excellent agreement between the @xmath19he and @xmath18he data was observed in the temperature range of overlap .
magnetic susceptibility measurements of polycrystalline band@xmath0o@xmath1 in an applied field @xmath20 @xmath4 0.1 t are presented in fig .
[ characterization ] ( plotted as 1/@xmath21 vs. @xmath6 ) and the results are in good agreement with previous work@xcite .
there was no observable difference for data collected under field - cooled and zero field - cooled conditions , and a broad valley is apparent in 1/@xmath21 around 2.2 k , likely corresponding to the onset of an antiferromagnetic transition .
the high temperature data is well - described by a curie - weiss law , with the value of the weiss temperature @xmath22 strongly dependent on the chosen fitting range , as first discussed in ref .
@xcite . fitting the data to the curie - weiss law
between 40 and 100 k yields @xmath22 @xmath4 -24 k and an effective moment
@xmath23 @xmath4 3.40(1 ) @xmath24 .
the effective moment is close to the expected value of 3.62 @xmath24 for nd@xmath25 .
neutron diffraction data from hb-2a with a neutron wavelength of 2.41 is depicted in fig .
[ neutron_diffraction ] at temperatures both above and below @xmath3 .
successful rietveld refinements were performed with the room temperature space group _ pnam _ , indicating that there are no structural phase transitions down to 300 mk .
the lattice constants at 300 mk refined as @xmath26 @xmath4 10.573(1 ) ,
@xmath27 @xmath4 12.432(1 ) , and @xmath2 @xmath4 3.601(1 ) .
several new bragg peaks appeared below @xmath3 , as shown in fig .
[ neutron_diffraction](b ) , and these could all be indexed with the propagation vector @xmath5 @xmath4 ( 0 1/2 1/2 ) .
the temperature dependence of the most intense magnetic peak ( 1 1/2 1/2 ) is plotted in fig .
[ neutron_diffraction](c ) , with the onset clearly observed around @xmath3 @xmath4 1.7 k , in good agreement with previous work@xcite .
diffuse scattering centered about @xmath28 @xmath4 ( 1 1/2 1/2 ) is also observed in the data , as shown most clearly in fig .
[ neutron_diffraction](d ) .
representational analysis allowed the possible magnetic structures to be constrained on the basis of the crystal symmetry .
there are two possible irreducible representations @xmath29 and @xmath30 ( kovalev s notation@xcite ) , each with six basis vectors , and therefore there are several possible candidates for the magnetic structure of band@xmath0o@xmath1 .
however , when refinements were attempted with several different models , the component of the nd magnetic moments along @xmath31 was consistently found to be negligible ( @xmath11 0.2 @xmath24 ) .
this parameter was assumed to be zero in subsequent refinements as a result .
recent work on the isostructural srr@xmath0o@xmath1 family has emphasized the important role that the different crystal field environments of the two r sites play in governing the magnetic properties of these materials@xcite .
more specifically , neutron diffraction has generally found that the inequivalent zigzag chains host different types of magnetic order or they are characterized by different ordered moment sizes . in some cases ,
only one of the two r sites was found to exhibit long - range order@xcite .
for these reasons , refinements were attempted with equal moments on the two sites and also with the moment on one r site constrained to be zero .
the best statistical refinements correspond to the scenario where only one nd site participates in the long - range ordering .
there are eight magnetic configurations that yield nearly identical fits of the data .
four of the models consist of collinear uudd intrachain spin arrangements , which are expected for classical ising @xmath8-@xmath7 chains in the large afm @xmath7 limit@xcite .
we note that the related systems srho@xmath0o@xmath1 and srdy@xmath0o@xmath1 have already been described as ising @xmath8-@xmath7 chain systems@xcite . for these reasons , the four intrachain collinear structures represent the most likely ordering configurations for band@xmath0o@xmath1 .
the other four possible models consist of non - collinear intrachain spin textures , with nnn spins along the chains afm - coupled .
these spin structures correspond to helices running along the chain direction for the special case when the nn spins are 90@xmath16 out of phase ( i.e. when four consecutive spins make up a complete period for the helix ) . while helical order is the expected ground state for an afm heisenberg @xmath8-@xmath7 model when @xmath7 @xmath32 0.25@xmath8@xcite , these non - collinear models only yield equivalent statistical refinements to the uudd models when the phases between consecutive pairs of nn spins are not constrained and refine to 63.6@xmath16 and 116.4@xmath16 .
the resulting structures are unlikely from a physical standpoint and therefore are discarded .
npd can not determine a unique magnetic structure for band@xmath0o@xmath1 because the refinements are insensitive to two different features of the system .
firstly , the r site that corresponds to the long - range order can not be distinguished by this technique .
this issue was first discussed for srer@xmath0o@xmath1@xcite .
the situation for the present case is very similar to the er system , as only one er site was found to show long - range order .
secondly , the relative phase along @xmath31 of the uudd ordered zigzag chains centered around @xmath33 @xmath4 0 and @xmath33 @xmath4 0.5 remains unknown .
equivalent refinements are obtained if the @xmath8 bonds within the chemical unit cell ( i.e. connecting nd atoms with @xmath34 @xmath4 0.25 and @xmath34 @xmath4 0.75 ) are fm and afm for the @xmath33 @xmath4 0 and @xmath33 @xmath4 0.5 chains respectively or vice versa . single crystal polarized neutron diffraction or point charge calculations of the crystal field levels may be able to help determine the type of long - range magnetic order in band@xmath0o@xmath1 unambiguously .
the four intrachain collinear structures are presented in fig .
[ mag_structure ] , with schematics shown for both the @xmath13 and @xmath35-planes in each case and the relevant exchange constants indicated in ( a ) .
all the possibilities have nd moments @xmath36 @xmath4 ( 1.39(1 ) , 2.25(1 ) , 0 ) @xmath24 .
the magnitude of the moment is 2.65(1 ) @xmath24 , which is somewhat lower than the expected value of 3.3 @xmath24 , but more reasonable than the value of 1.92 @xmath24 if one assumes that both nd sites are long - range ordered .
the reduced moment may be a consequence of strong frustration , as it is impossible to satisfy all the @xmath8 interactions within these spin configurations whether they are afm or fm .
the presence of the diffuse scattering below @xmath3 , as first discussed above , provides further evidence that only one nd site participates in the long - range order .
furthermore , since the strongest diffuse scattering feature is centered at the same position , @xmath28 @xmath4 ( 1 1/2 1/2 ) , as the most intense , resolution - limited , magnetic bragg peak , it is likely that the other type of nd spin chain exhibits quasi-1d , uudd magnetic order with the spins also confined to the @xmath13-plane . to further characterize band@xmath0o@xmath1 ,
the specific heat @xmath37 was examined as a function of applied fields @xmath20 .
the @xmath38 anomaly observed in zero field is suppressed with increasing field , as shown in fig .
[ cp](a ) . figure [ cp](b )
depicts @xmath39 data for @xmath20 @xmath32 3 t where the anomaly at the phase transition shows significant broadening .
we interpret this maximum as a schottky peak for @xmath20 @xmath40 4 t because it moves to higher temperatures with further increasing fields and it is found where our other measurements indicate that the afm order is suppressed .
furthermore , we estimate the entropy @xmath9
@xmath4 @xmath41 from the @xmath39 data , with the results shown in fig .
[ cp](c ) .
the jump @xmath42 at the phase transition in at all applied fields measured is less than 5.763 j / mol - k ( @xmath43 ) per individual nd atom . assuming a quasi - spin @xmath9 @xmath4 1/2 or even higher spin state for nd , this small entropy gain at the transitions point to incomplete magnetic ordering among the two nd magnetic sublattices .
magnetization measurements @xmath44 down to 0.46 k complement the neutron scattering and @xmath39 data .
representative @xmath45 vs. @xmath20 measurements are shown in fig .
[ mag](a ) with @xmath46 vs. @xmath20 shown in the inset , while @xmath21 ( i.e. @xmath47 ) vs. @xmath6 is depicted in ( b ) .
the @xmath48 curves in ( a ) show an s - shape that gets more pronounced with lower temperatures .
no full saturation is observed up to 7 t. instead , the maximum magnetization is only about 1.2 @xmath24 per nd .
this is well below the full moment of the free nd - atom and supports : ( i ) the results of the neutron scattering experiments and ( ii ) a smaller than @xmath10 @xmath4 9/2 spin state at low temperature , as found in the entropy analysis .
furthermore , two extrema are observed in @xmath46 at @xmath20 @xmath4 2.75 t and 4 t for @xmath6 @xmath4 0.46
k as shown in the inset of fig .
[ mag](a ) .
similar features from magnetization data have been used previously to identify field - induced phase transitions in other magnetic systems such as sr@xmath15cr@xmath0o@xmath49@xcite , although these extrema in @xmath46 only depict the transitions accurately as @xmath6 @xmath50 0@xcite .
a similar issue seems to be apparent in the present case , as the extrema are severely broadened with increasing @xmath6 .
the ordering transitions are also apparent in the @xmath21 vs. @xmath6 data , defined by a sharp drop in @xmath21 with decreasing @xmath6 .
this feature is completely suppressed above 4 t where the magnetic moments are fully polarized , as indicated by a constant @xmath21 behavior for @xmath6 @xmath50 0 . in low magnetic fields such as 0.1 t , a curie - weiss contribution to the susceptibility
is observed below 0.8 k , possibly arising from a small amount of magnetic impurity or due to the nd moments of band@xmath0o@xmath1 that do not participate in the long - range magnetic ordering .
however , the moments associated with this signal are fully polarized above 1 t and do not influence our experimental findings .
the most likely spin structures for the intermediate and high - field states correspond to an uud intrachain configuration and a paramagnetic phase respectively .
a field - induced uud phase has been predicted for classical ising @xmath8-@xmath7 chain systems with uudd magnetic structures in zero field@xcite , and therefore this possibility is in good agreement with conclusions made from the npd data .
we note that the 1/3 magnetization plateau corresponding to the uud state has already been found in single crystal data of srr@xmath0o@xmath1 ( r @xmath4 er , dy , ho)@xcite when @xmath20 is applied parallel to the spins within the uudd chains . in the present case of band@xmath0o@xmath1 ,
the uud state should give rise to a 1/3 magnetization plateau when @xmath20 is applied in the @xmath13-plane , but this feature is not expected when @xmath20 @xmath51 @xmath31 , and therefore the plateau likely gets smeared out in the polycrystalline data .
the combined magnetization and heat capacity data show that a modest magnetic field of only @xmath20 @xmath4 4 t is enough to drive the system to a completely paramagnetic state , as summarized in the phase diagram presented in fig .
[ phase_diagram ] .
the sharp @xmath38 anomalies observed in the @xmath39 measurements of our polycrystalline sample , combined with the neutron scattering and magnetization results discussed above , imply that the magnetic anisotropy of band@xmath0o@xmath1 is small but plays an important role in the low temperature magnetic properties .
the phase boundary between the low - field uudd state and the intermediate - field ordered state is not indicated on the phase diagram , as this boundary is expected to be orientation - dependent .
the growth of band@xmath0o@xmath1 single crystals would be highly beneficial for studying the @xmath6-@xmath20 phase diagrams along different crystalline directions , so the type of magnetic anisotropy and field - induced phases can be definitively established in this material .
we have investigated polycrystalline samples of the zigzag chain system band@xmath0o@xmath1 with a combination of magnetization , heat capacity , and neutron diffraction measurements .
neutron diffraction data shows evidence for magnetic bragg peaks below @xmath3 @xmath4 1.7 k that can be indexed with a propagation vector of @xmath5 @xmath4 ( 0 1/2 1/2 ) .
four candidate magnetic structures are consistent with the data , each arising from only one of the two types of nd zigzag chains , with the ordered spins lying in the @xmath13-plane .
all possible magnetic structures are composed of uudd spin chains , which is exactly the prediction for a classical ising @xmath8-@xmath7 chain in the large afm @xmath7 limit .
furthermore , low temperature magnetization and heat capacity measurements as a function of applied field reveal that the order can be completely suppressed in an applied field of only 4 t , with evidence found for an intermediate field - induced state .
one possibility is that the intermediate state corresponds to an uud spin structure , also predicted for the classical ising @xmath8-@xmath7 chain model .
future inelastic neutron scattering measurements and point charge calculations of the crystal field excitations for the systems in the ar@xmath0o@xmath1 family would help to clarify the anisotropy of the r atoms and lead to a better understanding of the magnetism in these frustrated magnets .
this research was supported by the us department of energy , office of basic energy sciences .
a.a.a . and v.o.g
were supported by the scientific user facilities division .
work at los alamos , j .- q.y . and
d.m . were supported by the materials science and engineering division .
the neutron experiments were performed at the high flux isotope reactor , which is sponsored by the scientific user facilities division .
gardner , m.j.p .
gingras and j.e .
greedan , rev .
82 * , 53 ( 2010 ) .
helton , k. matan , m.p .
shores , e.a .
nytko , b.m .
bartlett , y. yoshida , y. takano , a. suslov , y. qiu , j.h .
chung , d.g .
nocera , and y.s .
lee , phys .
* 98 * , 107204 ( 2007 ) .
t. aharen , j.e .
greedan , c.a .
bridges , a.a .
aczel , j. rodriguez , g.j .
macdougall , g.m .
luke , t. imai , v.k .
michaelis , s. kroeker , h.d .
zhou , c.r .
wiebe , and l.m.d .
cranswick , phys .
b * 81 * , 224409 ( 2010 ) .
t. aharen , j.e .
greedan , c.a .
bridges , a.a .
aczel , j. rodriguez , g.j .
macdougall , g.m .
luke , v.k .
michaelis , s. kroeker , c.r .
wiebe , h.d .
zhou , and l.m.d .
cranswick , phys .
b * 81 * , 064436 ( 2010 ) .
cheong and m. mostovoy , nature materials * 6 * , 13 ( 2007 ) .
silverstein , k. fritsch , f. flicker , a.m. hallas , j.s .
gardner , y. qiu , g. ehlers , a.t .
savici , z. yamani , k.a .
ross , b.d .
gaulin , m.j.p .
gingras , j.a.m .
paddison , k. foyevtsova , r. valenti , f. hawthorne , c.r .
wiebe , and h.d .
zhou , phys .
b * 89 * , 054433 ( 2014 ) .
gardner , s.r .
dunsiger , b.d .
gaulin , m.j.p .
gingras , j.e .
greedan , r.f .
kiefl , m.d .
lumsden , w.a .
macfarlane , n.p .
raju , j.e .
sonier , i. swainson , and z. tun , phys .
lett . * 82 * , 1012 ( 1999 ) .
gingras , b.c .
den hertog , m. faucher , j.s .
gardner , s.r .
dunsiger , l.j .
chang , b.d .
gaulin , n.p .
raju , and j.e .
greedan , phys .
b * 62 * , 6496 ( 2000 )
harris , s.t .
bramwell , d.f .
mcmorrow , t. zeiske and k.w .
godfrey , phys .
lett . * 79 * , 2554 ( 1997 ) .
ramirez , a. hayashi , r.j .
cava , r. siddharthan , and b.s .
shastry , nature * 399 * , 333 ( 1999 ) .
den hertog and m.j.p .
gingras , phys .
84 * , 3430 ( 2000 ) .
h. karunadasa , q. huang , b.g .
ueland , j.w .
lynn , p. schiffer , k.a .
regan , and r.j .
cava , phys .
b * 71 * , 144414 ( 2005 ) .
y. doi , w. nakamori , and y. hinatsu , j. phys .
matt . * 18 * , 333 ( 2006 ) .
igarashi and t. tonegawa , phys .
b * 40 * , 756 ( 1989 ) .
f. heidrich - meisner , i.a .
sergienko , a.e .
feiguin , and e.r .
dagotto , phys .
b * 75 * , 064413 ( 2007 ) .
a. fennell , v.y .
pomjakushin , a. uldry , b. delley , b. prevost , a. desilets - benoit , a.d .
bianchi , r.i .
bewley , b.r .
hansen , t. klimczuk , r.j .
cava , and m. kenzelmann , phys .
b. * 89 * , 224511 ( 2014 ) .
o.a . petrenko , g. balakrishnan , n.r .
wilson , s. de brion , e. suard , and l.c .
chapon , phys .
b * 78 * , 184410 ( 2008 ) .
hayes , g. balakrishnan , p.p .
deen , p. manuel , l.c .
chapon , and o.a .
petrenko , phys .
b * 84 * , 174435 ( 2011 ) .
o. young , l.c .
chapon and o.a .
petrenko , journal of physics : conference series * 391 * , 012081 ( 2012 ) .
o. young , a.r .
wildes , p. manuel , b. ouladdiaf , d.d .
khalyavin , g. balakrishnan , and o.a .
petrenko , phys .
b * 88 * , 024411 ( 2013 ) .
wen , w. tian , v.o .
garlea , s.m .
koohpayeh , t.m .
mcqueen , h .- f .
li , j .- q .
yan , d. vaknin , and c.l .
broholm , arxiv:1407.1341 ( unpublished ) .
quintero - castro , b. lake , m. reehuis , a. niazi , h. ryll , a.t.m.n .
islam , t. fennell , s.a.j .
kimber , b. klemke , j. ollivier , v.g .
sakai , p.p .
deen , and h. mutka , phys .
b * 86 * , 064203 ( 2012 ) .
h. li , b. hou , a. wildes , a. senyshyn , k. schmalzl , w. schmidt , c. zhang , t. bruckel , and g. roth , arxiv : 1404.0044 ( unpublished ) .
j. rodriguez - carvajal , physica b * 192 * , 55 ( 1993 ) .
a.s . wills , physica b * 276 * , 680 ( 2000 ) .
kovalev , _ representations of the crystallographic space groups edition 2 _ , ( gordon and breach publishers , switzerland , 1993 ) . h. de raedt and b. de raedt , phys .
b * 19 * , 2595 ( 1979 ) .
i. harada and h.j .
mikeska , z. phys .
72 * , 391 ( 1988 ) .
aczel , y. kohama , c. marcenat , f. weickert , m. jaime , o.e .
ayala - valenzuela , r.d .
mcdonald , s.d .
selesnic , h.a .
dabkowska , and g.m .
luke , phys .
103 * , 207203 ( 2009 ) a.a .
aczel , ph.d .
thesis , mcmaster university ( 2010 ) t.j .
hayes , o. young , g. balakrishnan and o.a .
petrenko , j. phys .
soc . jpn . *
81 * , 024708 ( 2012 ) . | the ar@xmath0o@xmath1 family ( r = rare earth ) have recently been attracting interest as a new series of frustrated magnets , with the magnetic r atoms forming zigzag chains running along the @xmath2-axis .
we have investigated polycrystalline band@xmath0o@xmath1 with a combination of magnetization , heat capacity , and neutron powder diffraction ( npd ) measurements .
magnetic bragg peaks are observed below
@xmath3 @xmath4 1.7 k , and they can be indexed with a propagation vector of @xmath5 @xmath4 ( 0 1/2 1/2 ) .
the signal from magnetic diffraction is well described by long - range ordering from only one of the two types of nd zigzag chains , with collinear up - up - down - down intrachain spin configurations .
furthermore , low temperature magnetization and heat capacity measurements reveal two field - induced spin transitions at 2.5 t and 4 t for @xmath6 @xmath4 0.46 k. the high field phase is paramagnetic , while the intermediate field state may arise from a spin transition of the long - range ordered nd chains , resulting in an up - up - down intrachain spin configuration .
the proposed intermediate field state is consistent with the magnetic structure determined in zero field for these chains by npd , as both phases are predicted for the classical ising chain model with nearest neighbor and next nearest neighbor interactions . |
astronomical observations brought important changes in modern cosmology @xcite .
while the type ia supernovae ( snia ) data are most often employed , the cosmic microwave background ( cmb ) observations and big - bang nucleosynthesis ( bbn ) analysis can also be used .
they allow the exotic physics of cosmological models to be checked against the observational data @xcite .
there is an increasing effort to develop some cosmological and astrophysical tools to search for new physics beyond the standard model .
the metric - affine gravity ( mag ) cosmological model with the robertson - walker symmetry was investigated by three different groups .
first , it was considered the model with triplet ansatz in vacuum @xcite .
second , it was considered the dilational hyperfluid model @xcite .
third , it was shown that on the level of the field equations the special case of the mag model is equivalent to a model in the weyl - cartan spacetime if we choose a model parameter in the special form ( @xmath2 ) @xcite .
moreover after redefinition of some variables the second and third approach gives the same set of dynamical equations .
the analysis of constraints on parameters in the mag model can be addressed in all three approach , but we adopt the last one proposed by puetzfeld and chen @xcite . note that in the model with dust matter on the brane , apart from dark radiation which scales like @xmath3 , there is a correction of the type @xmath1 to the einstein equations on the brane which arise from the influence of a bulk geometry @xcite .
the term scaling like @xmath1 also appears in the friedmann - robertson - walker model with spinning fluid @xcite .
it is possible to establish formally the one to one correspondence between the mag model and either the randall - sundrum brane model when positive values of the non - riemannian contribution to effective energy is admitted or the spinning fluid filled cosmology when this contribution is negative .
however , if one takes the pure randall - sundrum type model then there is a constraint on the brane tension parameter coming from the theory itself .
the brane tension parameter is not less than about ( @xmath4 gev)@xmath5 .
the mag model is free from such a theoretical constraint . in our further discussion
we examine the flat models which is motivated by the cmb wmap observations @xcite and consider the following formula for the friedmann first integral @xmath6 where @xmath7 is the hubble function , @xmath8 is the cosmological time , @xmath9 is the scale factor , @xmath10 , @xmath11 and @xmath0 are the density parameters for dust matter , the cosmological constant and fictitious fluid which mimics `` non - riemannian effects '' , respectively .
their values in the present epoch are marked by the index `` 0 '' .
all density parameters satisfy the constraint condition @xmath12 the density parameter for the fictitious fluid is defined as @xcite @xmath13 where @xmath14 the sign of the parameter @xmath15 is undetermined and it can assume both positive and negative values .
let us start from the reestimation of the models parameters by using the latest sample of snia data @xcite .
the motivation to study the sn constraint is to find the best estimation available from the latest data which gives the narrowest constraint for this method . in the next sections we compare it with constraints obtained from other methods .
riess et al.s sample contains 157 type ia supernovae @xcite .
we consider the flat model with and without priors on the @xmath0 and @xmath10 .
we assume that the former can be of any value or only nonnegative , and the latter is nonnegative or equal @xmath16 @xcite .
we estimate the best fits of the model parameters ( table [ tab:1 ] ) .
additionally we find the maximum likelihood estimates with the errors at @xmath17 level ( table [ tab:2 ] ) .
.best fit estimation of the model parameters from the snia data [ cols="<,^,^,^,^,^,^",options="header " , ] if we choose the @xmath18 km / s mpc then we obtain the agreement with the observation of the location of the first peak for three non - zero values of the parameter @xmath0 . as it is shown on figure [ fig:2 ]
there are two positive and one negative values of this parameter for which the mag model is admissible .
all these distinguished values of @xmath0 are in agreement with the result obtained from snia because the @xmath17 confidence interval for this parameter obtained from the snia data contains these three points . while the snia estimations give the possibility that @xmath0 is equal zero , the cmb calculations seem to exclude this case because the zero value of @xmath0 requires the first peak location at @xmath19 .
if we choose the @xmath20 km / s mpc than one of positive values of @xmath0 move to zero , while the second one move a little to the right .
we also calculated the age of the universe in the mag model .
we find that the difference in the age of the universe is smaller than 1 mln years for all three values of @xmath0 . assuming that @xmath21 the age of the universe is @xmath22 gyr for @xmath23 km / s mpc , and @xmath24 gyr for @xmath25 km / s mpc .
the globular cluster analysis indicated that the age of the universe is @xmath26 gyr @xcite .
.,title="fig:",scaledwidth=48.0% ] .,title="fig:",scaledwidth=48.0% ]
it is well known that the big - bang nucleosynthesis ( bbn ) is the very well tested area of cosmology and does not allow for any significant deviation from the standard expansion law apart from very early times ( i.e. , before the onset of bbn ) .
the prediction of standard bbn is in well agreement with observations of abundance of light elements .
therefore , all nonstandard terms added to the friedmann equation should give only negligible small modifications during the bbn epoch to have the nucleosynthesis process unchanged .
in our opinion the consistency with bbn is a crucial issue in the mag models where the nonstandard term @xmath1 in the friedmann equation is added ( see also discussion in @xcite ) .
this additional term scales like @xmath27 .
it is clear that such a term has either accelerated ( @xmath28 ) or decelerated ( @xmath29 ) impact on the universe expansion .
going backwards in time this term would become dominant at some redshift .
if it would happen before the bbn epoch , the radiation domination would never occur and the all bbn predictions would be lost .
the domination of the fictitious fluid @xmath30 should end before the bbn epoch starts otherwise the nucleosynthesis process would be dramatically modified .
if we assume that the bbn result are preserved in the mag models we obtain another constraint on the amount of @xmath0 .
let us assume that the model modification is negligible small during the bbn epoch and the nucleosynthesis process is unchanged .
it means that the contribution of the mag term @xmath0 can not dominate over the radiation term @xmath31 before the beginning of bbn ( @xmath32 ) @xmath33 the values of @xmath34 obtained as best fits in the snia data analysis as well as the smallest nonzero value of @xmath35 calculated in the cmb analysis are unrealistic in the light of the above result .
if we take into consideration the maximum likelihood analysis of snia data we have the possibility that the value of @xmath0 is lower than @xmath36 in the @xmath17 confidence interval . in the case of the cmb analysis only the value of the hubble constant close to @xmath37 km / s mpc gives the zero or close to zero value of @xmath0 .
the paper discusses observational constraint on `` energy contributions '' arising in certain cosmological models based on mag .
in particular it is focused on the nonstandard term @xmath1 .
we test this model against the snia data , the location of the peaks of the cmb power spectrum , and constraints from the bbn .
the mag model fits well to snia data and the estimations give the amount of fluid @xmath0 to be order of magnitude @xmath38 , and the hubble constant is close to @xmath39 km / s mpc .
let us note that these results are compatible with constraints from friib radio galaxies and x - ray gas mass fractions @xcite .
the cmb analysis gives that the hubble constant is @xmath37 km / s mpc which gives the too low age of the universe in comparison with the age of globular clusters .
taking lower value of the hubble constant obtained from snia estimation resolves the problem of the age .
however , the location of the first peak shifts to the right and is in conflict with the observed location .
the introducing of the non - riemannian structure of the underlying spacetime moves the location of the first peak back and this mag model agrees with the cmb observations .
the analysis of the cmb in this model can not distinguish the character of the fictitious fluid and we do not know whether the parameter @xmath0 is positive or negative .
the absolute values of @xmath0 obtained in the mag model from the cmb analysis with @xmath40 seems to be too large in comparison to the limit obtained from the bbn analysis . using the bbn analysis we pointed out that the mag part of the energy density to its present density parameter is of order @xmath41 .
the limit of this order leads to the conclusion that the mag model is virtually ruled out .
however , we must remember that we insist that the mag model does not change the physics during and after the bbn epoch . in this context , the merit of the snia analysis is its independency from any assumption on physical processes in the early universe . | we study metric - affine gravity ( mag ) inspired cosmological models .
those models were statistically estimated using the snia data .
we also use the cosmic microwave background observations and the big - bang nucleosynthesis analysis to constrain the density parameter @xmath0 which is related to the non - riemannian structure of the underlying spacetime .
we argue that while the models are statistically admissible from the snia analysis , complementary stricter limits obtained from the cmb and bbn indicate that the models with density parameters with a @xmath1 scaling behaviour are virtually ruled out .
if we assume the validity of the particular mag based cosmological model throughout all stages of the universe , the parameter estimates from the cmb and bbn put a stronger limit , in comparison to the snia data , on the presence of non - riemannian structures at low redshifts . |
ls i @xmath061@xmath1303 is one of a handful high - mass x - ray binaries that have been detected at all frequencies , including tev and gev energies .
its nature is still under debate , with rotationally - powered pulsar - composed systems ( see maraschi & treves 1981 , dubus 2006 ) and microquasar jets ( see bosch - ramon & khangulyan 2009 for a review ) being discussed .
recently , evidence favoring ls i @xmath061@xmath1303as the source of a very short x - ray burst led to the analysis of a third alternative , in which ls i @xmath061@xmath1303 is a magnetar binary ( see torres et al .
2011 , also bednarek 2009 , and dubus 2010 ) .
long - term monitoring of the source , ideally across all wavelengths , is a key ingredient to disentangle differences in behavior which could point to the underlying source nature . here
, we report on the analysis of _ rxte_-proportional counter array ( pca ) monitoring observations of the @xmath3-ray binary system ls i @xmath061@xmath1303 .
the dataset we consider covers the period between 2007 september and 2011 september ( 2007 - 08 - 28 2011 - 09 - 15 ) , and it is the longest monitoring campaign done for this source .
smaller datasets included in this campaign have been analyzed and results have been presented in our previous papers i and ii ( torres et al .
2010 , and li et al .
2011 , respectively ) . in them
, we focused on establishing the orbit - to - orbit x - ray variability and on studying the spectral properties and the flares .
the current data enhances in more than one additional year the reported coverage on the source , and for the first time , it is sufficient to consider the possible super - orbital modulation of the x - ray emission from ls i @xmath061@xmath1303 .
our current analysis includes 473 _ rxte_-pca pointed observations identified by proposal numbers 93100 , 93101 , 93102 , 93017 , 94102 , 95102 and 96102 .
the analysis of pca data was performed using heasoft 6.9 .
we filtered the data using the standard _ rxte_-pca criteria . only pcu2 ( in the 0 - 4 numbering scheme ) has been used for the analysis because it was the only proportional counter unit ( pcu ) that was 100% on during the observations .
we select time intervals where the source elevation is @xmath410@xmath1 and the pointing offset is @xmath5 .
the background file used in the analysis of pca data is the most recent available one from the heasarc web site for faint sources , and detector breakdown events have been removed .
our flux and count rate values are given for an energy range of 330 kev .
at first , we consider the complete x - ray lightcurve of our campaign , and fitted it with a constant to obtain an average flux , resulting in ( [email protected] ) counts s@xmath6 . in order to remove the influence of the several ks - long flares detected from the source
, we cut all observations that presented a count rate larger than three times this average .
the remaining on - source time amounts to 684.3 ks ( 460 observations ) , and it is uniformly distributed in the system s orbital ( between 60 and 80 ks per each 0.1 of orbital phase bin ) as well as in the system s super - orbital phase ( around 60 ks per each 0.1 of super - orbital phase bin , except at super - orbital phase 0.8 , where an intensive campaign increased the exposure to 120 ks ) as defined by the radio observations ( gregory 2002 ) . given a 6-months time bin ( or approximately 6.8 orbits of the system )
, we take the peak x - ray flux in that period and compute the modulated flux fraction .
the latter is defined as ( @xmath7 ) , where @xmath8 and @xmath9 are the maximum and minimum count rates in the 330 kev orbital profile of that period , after background subtraction .
the minimum count rates are roughly constant around 1 count s@xmath6 all along the observation time .
results are shown in figure [ super - orb ] .
table 1 presents the values of the reduced @xmath10 for fitting different models to the modulation fraction and the peak flux in x - rays .
it compares the results of fitting a horizontal line , a linear fit , and two sinusoidal functions .
one of the latter has the same period and phase of the radio modulation ( from gregory 2002 , labelled as ` radio ' in table 1 , dotted line in figure 1 ) .
the other sine function has the same period as in radio but allowing for a phase shift from it ( a solid line in figure 1 , labelled as ` shifted ' in table 1 ) .
it is clear that there is variability in the data ( and thus that a constant fit is unacceptable ) , as well as that the sinusoidal description with a phase - shift is better than the linear one , which is obvious by visual inspection of figure [ super - orb ] .
the phase shift derived by fitting the modulated fraction is 281.8 @xmath2 44.6 days , corresponding in phase to @xmath11 of the 1667@xmath28 days super - orbital period .
the phase shift derived by fitting instead the maximum flux is 300.1 @xmath2 39.1 days , and results compatible with the former .
[ tab1 ] .reduced @xmath10 for fitting different models to the modulation fraction and the peak flux in x - rays . [ cols="<,<,<,<,<",options="header " , ] figure [ scales ] analyzes whether there is anything special concerning the 6 months time bin just chosen to present the former results .
it presents the peak flux and modulated fraction in different time bins , from 4 to 12 months .
the black lines in figure [ scales ] shows the sine fitting with a 1667 days period fixed .
the black lines in all left ( right ) panels are the same as the one depicted in the left ( right ) panel of figure [ super - orb ] .
it is interesting to note that the smaller the time bin the larger it is the scattering around the sinusoidal fit , which can be understood as an effect of the increasing similarity between the time bin itself and the orbital period of the system ( of [email protected] , gregory 2002 ) .
indeed , orbit - to - orbit variability is known to exist in our data , and can be similar or larger than the super - orbital induced variability at times ( e.g. , see the variations found for the same phase - bin in contiguous orbits in figure 4 of torres et al .
thus , the shorter the time bin , the less likely it is that the super - orbital induced variation could be detected , which may be sub - dominant to the local - in - time changes . on the contrary , as soon as the integrated time bin is large enough in comparison with the orbital period of ls i @xmath061@xmath1303 , the super - orbital variability is consistently observable .
this same fact makes a direct comparison of the sinusoidal trend of the super - orbital x - ray modulation of figure [ super - orb ] with the flux obtained from ls i @xmath061@xmath1303under short and isolated observations untrustable .
for example the observations at soft x - rays conducted by _ xmm - newton _
( neronov & chernyakova 2006 ; chernyakova et al .
2006 , sidoli et al .
2006 ) , _ chandra _ ( paredes et al .
2007 , rea et al . 2010 ) , _ asca _ ( leahy et al .
1997 ) , _ rosat _ ( goldoni & mereghetti 1995 ; taylor et al . 1996 ) , and _ einstein _ ( bignami et al . 1981 ) were all too short to cover even a single full orbit .
similarly , earlier campaigns with pca during the month of march 1996 ( harrison et al .
2000 , greiner and rau 2001 ) are also too short .
comparing their obtained flux values with the super - orbital trend would be meaningless since there is not enough time coverage to average out the possible orbit - to - orbit variations .
finally , we note that we have also analyzed 15 years of
_ rxte_-asm data on ls i @xmath061@xmath1303 , but the larger error bars on the count rates preclude to obtain any conclusion from that dataset .
we notice that the x - ray long - term monitoring ( 20072011 ) on ls i @xmath061@xmath1303 started about 7 years later from the end of the campaign used to determine the super - orbital period in radio ( 19772000 , see gregory 2002 ) .
we will assume then that the radio - determined super - orbital modulation of the source is , although possibly variable , active today with similar features as the ones claimed a decade earlier .
this appears possible given that recent reports of variation in the orbital radio maxima are of only @xmath120.1 in phase ( see trushkin & nizhelskij 2010 ) .
under such assumptions , we showed that there is a phase shift between the radio and the x - ray super orbital modulation .
interestingly , this shift is the same as the one hinted at between the radio and the h@xmath13 line ( zamanov et al .
1999 , zamanov & mart 2000 ) .
indeed , the optical observations that covered the period 19891999 were fitted with a period of @xmath121584 days ( zamanov et al .
1999 ) , a value reported prior to the work by gregory ( 2002 ) , where the super - orbital period was refined to 1667@xmath28 days . to investigate further the long - term modulation of ls i @xmath061@xmath1303 at optical wavelengths and how it compares with the current findings in x - rays
, we took the h@xmath13 data from table 4 of paredes et al .
( 1994 ) , table 1 of zamanov et al .
( 1999 ) , and figure 1 of zamanov et al .
( 2000 ) , and as stated in zamanov et al .
( 1999 ) , considered an error of 10% for all the equivalent widths .
we performed an analysis similar to the one done for x - rays in the previous section , and derived and optical phase lag of [email protected] days with respect to the radio phase .
thus , the optical phase lag is coincident with the one derived at x - rays , although the observations at the two bands are about 8 years apart .
the stellar disk of be stars are well known to grow larger as the equivalent width of the h@xmath13 emission line increases ( e.g. , hanushik et al .
1988 ; grundstrom & gies 2006 ) .
the optical variability has been most likely attributed to the cyclic variation of be circumstellar disk .
thus , the possible coincidence with the x - ray phase lag suggests that the stellar disk may play an important role also for the x - ray emission , and probably for the higher - energy non - thermal emission of ls i @xmath061@xmath1303 too .
the coincidence between the x - ray and optical shift with respect to the 1667 days radio modulation has to be taken with the necessary prudence prompted by it being based on non - simultaneous observations . in particular , it seems that the optical observations present the largest degree of variation in time .
we checked that in addition of the h@xmath13 measurements mentioned above , there are more recent ones in the works of e.g. , liu et al .
( 2005 ) , grundstrom et al .
( 2007 ) , zamanov et al .
( 2007 ) , and mcswain et al .
however , the latter span 0.51 ( at best , being usually much shorter ) of the super - orbital period , and as such we can not directly use them for a comparison in long - terms . nevertheless , they seem to hint at that the h@xmath13 variability is not strictly periodic or at least at a changing amplitude . based ( among other reasons discussed in torres et al .
2011 ) on the analysis of a the _ swift_-bat detection of a short , magnetar - like burst from the direction of ls i @xmath061@xmath1303 , we have proposed that the system s compact object is a high magnetic field , slow period pulsar . in that case
, we proved that the ls i @xmath061@xmath1303-system would most likely be subject to a flip - flop behavior , from a rotationally powered regime ( in apastron , also known as ejector ) , to a propeller regime ( in periastron ) along each of the system s eccentric orbits .
the multi - wavelength phenomenology can be put in the context of the former model , and in particular , also the highest energy tev emission , which has also shown low and high states which are apparently modulated by the same super - orbital period as well . within this model
, we notice that an increase in the accreted mass onto the compact object ( unavoidably linked to the mass - loss rate of the star ) by a factor of a few61@xmath1303 are given as the ratio between maximal and minimal values obtained either from radio emission ( a factor of 4 was determined by gregory & neish 2002 ) or from h@xmath13 measurements , which span from a factor of 5.6 ( gregory et al .
1989 ) to 1.5 ( zamanov et al . 1999 ) . ]
can put the system in a permanent propeller stage along the orbit , including at the apastron region .
this change of behavior for such an small change in mass loss rate can be the reason behind the evolution of the modulated fraction .
indeed , using the formulae in torres et al .
( 2011 ) , and considering to simplify that the condition @xmath14 , where @xmath15 stands for the magnetic radius and @xmath16 for the light cylinder , establishes both the out - of - ejector and into - ejector condition , one sees that the period mass - loss magnetic field relation for the apastron of ls
i @xmath061@xmath1303 is @xmath17 where @xmath18 represents a constant of order 1 , and we have assumed an eccentricity of 0.6 and a semi - major axis of @xmath19 cm . for periods shorter than the former , the system is in an ejector phase . for larger periods , it is in a propeller stage ( see torres et al .
2011 for details ) .
high values of magnetic field and slow periods would make the transition possible : a cyclical change by a factor of a few in @xmath20 can make the system to abandon the ejector phase in apastron . for instance , under a variation by a factor of 4 in @xmath20 , a case that leads to a super - orbital induced transition is given by a magnetic field of @xmath21 g , and a typical period of magnetars ( @xmath22 s ) .
this may also happen for smaller values of the magnetic field but only in the case of a relatively long period .
for instance , again under a variation by a factor of 4 in @xmath20 and for @xmath23 g , the period should be between 700 ms and 1s in order for the system to flip - flop in the super - orbital evolution , although no known pulsar in these parameter ranges has a rotational energy in excess of @xmath24 erg s@xmath6 ( atnf catalogue version : 1.43 ) , which would be needed to account for the multi - wavelength output of the system .
note in particular that the behavior of the ls i @xmath061@xmath1303 system containing a pulsar with @xmath25 g and @xmath26 ms would be unaffected by the cyclical variation of the mass - loss rate : it would act as an ejector in apastron along the whole super - orbital period .
the flip - flop mechanism can then be used to qualitatively explain why ls i @xmath061@xmath1303 has entered in a low tev state ( see , e.g. , acciari et al .
2010 ) when at the maximum of the radio super - orbital variability , but perhaps also to explain why the modulated x - ray flux fraction varies as we found in figure [ super - orb ] .
when the mass - loss rate is low , the inter - wind shock formed at the collision region between the pulsar and the stellar wind would be present at the broad apastron region ( and so will the tev emission there ) , disappearing at periastron .
in this situation , there are two contributors to the x - ray emission along the orbit , expected to be roughly at the same level ( e.g. , zamanov et al .
1999 ) ; the shock at apastron and the propeller at periastron , and the modulated fraction is consistently low . when at the maximum of the mass - loss rate , the inter - wind shock may not form , and abundant tev particles would not be produced since shocks at the magnetic radius are unable to reach tev energies .
thus there is only one process generating the x - ray emission along the system s orbit , the propeller , and the modulated fraction is then maximum . the exact position of the x - ray maximum along each of the orbits would depend on the local - in - time conditions of the accreted mass onto the compact object , which established the relative weight of the two x - ray contributors , and it is thus expected to vary beyond the super - orbital trend in short timescales , and not always be located at periastron .
however , given that the h@xmath13 cycle represents the cyclical modulation of the mass loss rate , it would be natural to expect that the x - ray emission be correlated with it in long timescales ( i.e. , with how much mass is falling towards the compact object , e.g. , see bednarek 2009 or bednarek & pabich 2011 ) .
zamanov et al .
( 2001 ) already discussed when the radio emission is expected to peak in each of the system orbits : the switch on of the ejector phase will activate the pulsar wind , creating a cavern around the neutron star which will start to expand .
this means that the radio outburst will peak with some delay after the change of regimes , which is supposed to happen somewhere after the periastron , when the accretion rate onto the compact object diminishes enough . in a cyclical variability of the mass loss rate of the star ,
the ejector - propeller transition moves in phase : at lower mass loss rates , the ejector will switch on earlier , and the radio outburst will peak at earlier orbital phases than at higher mass loss rates .
a generic tev and radio anti - correlation is thus expected since the more mass fuels the propeller phase the more violent the radio outburst will be , and the less effectively the inter - wind shock will generate tev particles .
the colored boxes in figure [ super - orb ] represent the times of the tev observations that covered the broadly - defined apastron region ( from albert et al .
2006 , 2008 , anderhub et al .
2009 , aleksic et al . 2011 ; acciari et al . 2008 , 2009 , 2010 ) .
those boxes colored in green denote the times for which the observations led only to imposing an upper - limit or to a detection with a flux that is about 3 times less than the one obtained at the discovery observations of 2006 ( albert et al . 2006 ) , which defines the low state .
the yellow boxes stand for those observations for which the level of the tev emission was roughly compatible with the original discovery .
there is a trend for finding a low tev state towards the maximum of the super - orbital low - frequency cycles .
this is perhaps more clearly seen in figure [ scales2 ] , where we plot the peak flux per orbit in tev ( all of them happening in the 0.61.0 orbital phase range ) as a function of superorbital phase , together with radio and h@xmath13 data
. however , the scarcity ( and non - simultaneity ) of the tev coverage precludes reaching a definite conclusion on whether there is an anti - correlation of the tev emission with the radio or with the h@xmath13 curves .
it would seem , however , that the tev emission is rather anti - correlated with the radio flux and not with h@xmath13 , but this could not be quantitatively proven with the data at hand , especially given the caveats of dealing with non - contemporaneous observations .
a simultaneous optical - tev campaign is needed to establish the nature of the anti - correlation .
the latter would be particularly useful for the forthcoming extrapolated radio maximum around october - november 2012 .
we acknowledge support from the grants aya2009 - 07391 and sgr2009 - 811 , as well as the formosa program tw2010005 , by the national natural science foundation of china via nsfc-10325313 , 10521001 , 10733010,11073021 , and 10821061 , the cas key project kjcx2-yw - t03 , and 973 program 2009cb824800 .
ypc thanks the natural science foundation of china for support via nsfc-11103020 and 11133002 .
nr is supported by a ramon y cajal fellowship .
we also acknowledge the use of the high energy astrophysics science archive research center ( heasarc ) , provided by nasa s goddard space flight center .
jl acknowledges the hospitality of ieec - csic , where this research was conducted .
+ albert j. , et al .
2006 , science , 312 , 1771 albert j. et al .
2008 , apj 684 , 1351 anderhub . ,
2009 , apj 693 , 303 aleksic j. , et al .
2011 , submitted .
acciari v. a. et al .
2008 , apj 679 , 1427 acciari v. a. et al .
2009 , apj 700 , 1034 acciari v. a. et al .
2010 , apj in press , arxiv:1105.0449 bednarek , w. 2009 , mnras 397 , 1420 bednarek , w. & pabich , j. 2011 , mnras 411 , 1701 bignami g.f.b . ,
1981 , apj 247 , l85 bosch - ramon v. , & khangulyan d. 2009 , international journal of modern physics d18 , 347 gregory p.c .
, huang - jian xu , bachhouse c.j . , & reid a. , 1989 , apj 339 , 1054 gregory p. c. 2002 , apj 575 , 427 gregory , p. c. , & neish , c. , 2002 , apj , 580 , 1133 greiner , j. & rau , a. , 2001 , a&a , 375 , 145 taylor a.r .
, et al . 1996 , a&a 305 , 817 torres d. f. , et al .
2010 , apj letters 719 , 104 ( paper i ) torres d. f. , et al .
2011 , apj in press , astro - ph ( arxiv:1109.5008 ) trushkin , s. & nizhelskij , n. 2010 , poster at the 38th cospar scientific assembly 2010 , accretion on compact objects and fast phenomena in multiwavelength era ( e13 ) , symposium e , session 13 , paper number e13 - 0075 - 10 ( poster , nr .
wed-152 ) zamanov r. k. , mart , j. , paredes j. m. et al . 1999 ,
a&a 351 , 543 zamanov r. k. , & mart j. m. , 2000 , a&a 358 , 55 zamanov r. k. , mart j. m. , & marziani p. 2001 , arxiv 0110114 zamanov r. k. , stoyanov k. a. , & tomov n. a. 2007 , commission 27 and 42 of the iau , information bulletin on variable stars , number 5776 | from the longest monitoring of ls i @xmath061@xmath1303 done to date by the _ rossi x - ray timing explorer _ ( _ rxte _ ) we found evidence for the long - sought , years - long modulation in the x - ray emission of the source . the time evolution of the modulated fraction in the orbital lightcurves can be well fitted with a sinusoidal function having a super - orbital period of 1667 days , the same as the one reported in non - contemporaneous radio measurements .
however , we have found a 281.8 @xmath2 44.6 days shift between the super - orbital variability found at radio frequencies extrapolated to the observation time of our campaign and what we found in the super - orbital modulation of the modulated fraction of our x - ray data .
we also find a super - orbital modulation in the maximum count rate of the orbital lightcurves , compatible with the former results , including the shift . |
in many - body quantum mechanics , the notion of particle indistinguishability and statistics plays a fundamental role .
namely , particles of the same kind are typically logically identical and fall into two classes : bosons or fermions , giving rise to such diverse phenomena as bose
einstein condensation in the former case , and the fermi sea with its implications for conduction bands , atomic structure , etc .
, in the latter .
however , while these are the only two options for fundamental particles that propagate in three - dimensional space , for quantum systems confined to lower dimensions there is a possibility for effective particles ( quasiparticles ) escaping the usual boson / fermion dichotomy .
we shall here consider the two - dimensional case where the quantum state of a system of @xmath0 particles at positions @xmath1 may be described by a square - integrable normalized wave function - valued wave functions corresponding to abelian anyons , while @xmath2-valued , possibly non - abelian , anyons are also possible @xcite . ]
@xmath3 , where @xmath4 is the probability density of finding the particles at positions @xmath5 . if the particles are indistinguishable the density needs to be symmetric : @xmath6 however , the exact phase of @xmath7 is not an observable quantity and therefore leaves room for an exchange phase , @xmath8 where @xmath9 ( @xmath10-periodic ) is called the statistics parameter .
logical consistency demands this to be independent of the considered pair if we by mean that we make a continuous counterclockwise simple exhange of this pair of particles . if @xmath11 the particles are bosons ( symmetric @xmath7 ) , and if @xmath12 they are fermions ( antisymmetric @xmath7 ) . because of a topological obstruction unique to two dimensions , it turns out that _ any _ phase @xmath13 is allowed @xcite ( see also @xcite ) , and the corresponding particles are therefore called anyons @xcite .
we refer to @xcite for extensive reviews on this topic .
the relative change of phase of a wave function @xmath7 with respect to changes of the coordinates , geometrically understood as due to the curvature of the corresponding complex line bundle of which @xmath7 is a section , is naturally described by a magnetic field , and also in the case of anyons one may model the above statistics phase as induced by a magnetic field .
namely , one could start with @xmath14 ( or @xmath15 ) being bosonic ( fermionic ) and then attach magnetic fluxes to the particles so that their winding around each other gives rise to the correct phase .
this is commonly called the magnetic gauge picture for anyons , and it is actually in this form that they may most realistically arise in a real physical system .
the most promising such realization is in the context of the fractional quantum hall effect ( fqhe ) @xcite , a strongly correlated planar electron ( or bosonic atom @xcite ) system in a strong transverse external magnetic field , where particles have the freedom to bind magnetic flux and thereby become anyons @xcite .
however , in this case the flux typically has some extent determined by the experimental conditions , and one therefore talks about extended anyons @xcite as opposed to the purely theoretical ( but conceptually attractive ) ideal anyons which are purely pointlike . denoting the size of the flux ,
say its radius if disk - shaped , by @xmath16 we can thus talk about @xmath17-extended anyons , and one may also introduce a dimensionless parameter @xmath18 to describe the state of the system , where @xmath19 denotes the average density of the particles .
the parameter @xmath20 is the ratio of the magnetic dimension to the average interparticle distance and has therefore been called the magnetic filling ratio in @xcite .
our interest in this paper is to study a free gas of such extended anyons , i.e. ignoring any additional interactions , and focusing on the most basic aspect : its ground - state energy . in the ideal non - interacting case ,
the quantum gas consisting of a large number of bosons or fermions in a large volume at fixed density has been completely understood since the early days of quantum mechanics and is nowadays often given as a text - book exercise .
however , the purely anyonic case @xmath21 still remains an unsolved problem after almost four decades , owing to the fact that the statistical many - body interaction can not be completely removed in favor of a pure one - body state - counting description .
the case of two anyons can be solved exactly @xcite , that of three and four anyons has been studied numerically @xcite , and beyond that various approximative descriptions have been proposed @xcite .
one of these is called average - field theory ( cf .
mean - field theory @xcite ) whereby the flux of the anyons is seen as sufficiently spread out ( in other words @xmath20 should be sufficiently large ) so that they are effectively moving in a ( locally ) uniform magnetic field , say @xmath22 where @xmath23 is the flux of each anyon and @xmath24 the local density , and therefore have a definite magnetic ground - state energy given by that of the lowest landau level , hence proportional to @xmath25 .
another approximation has been to assume that the gas is so dilute that only two - particle interactions are relevant @xcite . except for a small number of results concerning the mathematical formulation of the many - anyon problem @xcite , there has not been much progress on the rigorous mathematical side until recently . in @xcite the case of ideal anyons was considered using a local approach involving a relative magnetic hardy inequality and a local exclusion principle , leading to a first set of non - trivial rigorous bounds for the ground - state energy of the ideal anyon gas .
these bounds have an interesting non - trivial dependence on the statistics parameter @xmath26 in that they depend , in the many - body limit , solely on the quantity @xmath27 which is zero unless @xmath26 is an odd - numerator fraction @xmath28 ( reduced , with @xmath29 ) and in which case @xmath30 . in @xcite a fundamental question concerning operator domains for ideal anyons was settled and applications of the local energy bounds to interacting systems were considered .
also , the validity of the average - field approximation for the case of almost - bosonic ( @xmath31 ) @xmath17-extended anyons was proved in @xcite .
here we shall consider the homogeneous @xmath17-extended anyon gas and build on the local approach of @xcite to prove a lower bound for the ground - state energy per particle with statistics parameter @xmath32 and magnetic filling ratio @xmath33 of the form @xmath34 where @xmath35 is a universal constant and @xmath36 2\pi|\alpha| , & \quad \gamma \gtrsim 1 .
\end{array}\right.\ ] ] here @xmath37 denotes the first positive zero of the derivative of the bessel function @xmath38 ( and @xmath39 ) .
this bound effectively interpolates between a dilute regime involving and a high - density regime with a dependence on @xmath26 that matches that of average - field theory .
also in the case of even - numerator @xmath26 , where @xmath40 , the bound is strictly positive but vanishes in the dilute limit in a way similar to that of a dilute bose gas in two dimensions @xcite . this may however not be so surprising in the case that @xmath41 ( composite bosons ; cf .
@xcite ) , considering the periodicity for ideal anyons . in order to state our results
precisely we need to introduce some notation that will be used throughout the paper .
we take as our concrete model for @xmath17-extended anyons a set of @xmath0 identical charged bosons , to each of which has been attached a magnetic field in the shape of a disk with radius @xmath17 and total flux @xmath23 , and which is felt by all the other particles ( cf .
@xcite ) . such flux centered at the origin
can be given concretely by the magnetic vector potential @xmath42 with @xmath43 where @xmath44 , @xmath45 denotes the open ball / disk of radius @xmath17 centered at @xmath46 , and @xmath47 can be interpreted as a regularized distance .
starting from a conventional magnetic hamiltonian formulation , the ( non - relativistic ) free kinetic energy operator is then @xmath48 where we have normalized physical units so that @xmath49 and the magnetically coupled momentum operator for each particle @xmath50 is given by @xmath51 where @xmath52 corresponding to the total magnetic field felt by the particle @xmath53 @xmath54 we note that this form for the magnetic interaction is not only convenient but also realistic from the perspective of the fqhe @xcite .
also note that we allow for any @xmath9 here .
the operator acts on the bosonic space @xmath55 as an unbounded operator .
let us denote by @xmath56 the natural ( minimal as well as maximal ( * ? ? ?
* theorem 5 ) ) domain of @xmath57 then this is also the natural form domain of , and @xmath58 . in the case
@xmath59 ( as well as for @xmath11 ) we have @xmath60 , since @xmath61 is then a bounded perturbation of @xmath62 . on the other hand ,
if @xmath63 ( and @xmath64 ) these spaces are typically different ( see ( * ? ? ?
* section 2.2 ) ) . for @xmath63 and @xmath41 ( respectively @xmath65 ) , however , @xmath66 is gauge - equivalent to @xmath67 ( respectively @xmath68 ) : @xmath69 where @xmath70 is the isometry ( singular gauge transformation ) @xmath71 with @xmath72 the complex coordinate representatives of @xmath53 given by identifying @xmath73 with @xmath74 .
we define the one - body density associated with any normalized state @xmath75 by @xmath76 while @xmath77 denotes the average density if confined to a domain ( typically a square ) @xmath78 , i.e. @xmath79 .
furthermore , with @xmath80 the fat diagonal of the configuration space @xmath81 , we note that we may use the density of @xmath82 in the domain @xmath56 ( again , see ( * ? ? ?
* theorem 5 ) ) .
we are now ready to state our main results for @xmath17-extended anyons . for ideal anyons
the singular magnetic potential @xmath61 effectively excludes the diagonals @xmath83 from the configuration space , much like a strong repulsive point interaction . for @xmath17-extended anyons
we have instead the following effective repulsive short - range interaction of soft - disk type : [ lem : short - range ] + for any @xmath9 , @xmath59 , @xmath84 , and @xmath85 we have that @xmath86 now , defining ( denoted @xmath87 in @xcite ) @xmath88 let us recall the following local many - particle magnetic hardy inequality for ideal anyons which was given in ( * ? ? ?
* theorem 4 ) : [ thm : ideal - hardy ] let @xmath9 , @xmath89 , @xmath84 and @xmath90 be open and convex . then , for any @xmath91 , @xmath92 with the reduced support @xmath93 , and @xmath94 pairwise coordinates and distances .
our first main result is the following improvement and extension of the above inequality to @xmath17-extended anyons , thereby providing us with a concrete ( and indeed useful ) measure of the long - range effect of the statistical magnetic interaction : [ thm : long - range ] + let @xmath9 , @xmath16 , @xmath95 and @xmath90 be open and convex .
then , for any @xmath96 and @xmath97 , @xmath98 where @xmath99 may depend on the positions of all @xmath0 particles @xmath100 , @xmath101 describes a maximal annulus contained in @xmath102 ( with some @xmath17-dependent margins ) in terms of the relative coordinate , and @xmath103 for @xmath104 and @xmath105 is the square root of the smallest positive solution @xmath106 associated with the bessel equation @xmath107 on the interval @xmath108 $ ] with neumann boundary conditions , while @xmath109 for @xmath110 . in the non - extended case @xmath89 the inequality is valid with @xmath111 ( hence take @xmath112 ) , while for any @xmath113 it holds at least for @xmath114 . moreover , the function @xmath115 has the following properties : @xmath116 where @xmath37 denotes the first positive zero of the derivative of the bessel function @xmath38 ( and @xmath39 ) .
we note that the above inequality may in some sense be viewed as a refinement ( with respect to the relative angular part ) of the usual ( pointwise ) diamagnetic inequality : [ lem : diamagnetic ] + for any @xmath9 , @xmath16 , @xmath84 and @xmath96 we have that @xmath117 for @xmath118 , the vector potential satisfies @xmath119 and hence it is covered by standard theorems ; see e.g. ( * ? ? ?
* theorem 7.21 ) .
for @xmath89 it is not , but the above diamagnetic inequality still holds in this case , as was proved in ( * ? ? ?
* lemma 4 ) ( and actually our understanding of the form domain @xmath66 alluded to above depends on this general formulation of the inequality ) .
the diamagnetic inequality of lemma [ lem : diamagnetic ] says that the energy for anyons is always higher than that for free bosons , while the short - range inequality of lemma [ lem : short - range ] tells us that the anyons also feel an effective repulsion proportional to @xmath120 whenever they overlap .
taking a combination of these two bounds would then correspond to a two - dimensional soft - disk repulsive bose gas , whose energy in the dilute limit tends to zero logarithmically with the density ( here the magnetic filling ratio @xmath121 ) @xcite .
on the other hand , theorem [ thm : long - range ] provides a local bound for the energy in the form of a long - range inverse - square repulsion , whose strength depends on the fractionality of @xmath26 via @xmath122 .
while this ` statistical repulsion ' does not change the above repulsive picture much in the regime of high densities ( @xmath123 ) where the anyons already feel each other s magnetic fields by ( partially ) overlapping , it makes a significant difference in the dilute limit , resulting in a uniform bound for the energy from below in terms of @xmath124 . as in @xcite ,
the reason for the dependence on @xmath125 and not directly @xmath26 in this bound is the local gauge invariance of the pairwise relative magnetic potential .
namely , any even multiple of a unit flux enclosed in a two - particle exchange loop ( enclosing @xmath126 other particles in such a loop contributes @xmath127 multiples of an anyon flux in ) may be compensated for by an opposite and equally large orbital angular momentum of that same particle pair .
however , for odd - numerator rational @xmath26 there can never be a complete cancellation of this type , and therefore there is always some long - range pair repulsion , @xmath128 .
these effects are summarized in the following theorem concerning the @xmath17-extended anyon gas , which is our second main result : [ thm : intro - gas ] + let @xmath129 , where @xmath130 , denote the ground - state energy per particle and unit density of the extended anyon gas in the thermodynamic limit at fixed @xmath9 , @xmath16 and density @xmath131 where dirichlet boundary conditions have been imposed , that is @xmath132 then @xmath133 for some universal constants @xmath134 , @xmath135 ( is defined in lemma [ lem : localexclusionsr ] ) , and @xmath115 as in theorem [ thm : long - range ] .
furthermore , for any @xmath9 we have for the ideal anyon gas that @xmath136 at ( 0,0 ) plotted as a function of @xmath20 for @xmath137 and @xmath138 ( blue , yellow , green , red and purple ) in the hypothetical case @xmath139 , @xmath140 for illustrative purposes .
the figure to the left shows the general behavior over the full range of @xmath20 , while that on the right shows the behavior in the dilute limit plotted in logarithmic scale where the long - range dependence on @xmath125 becomes relevant.,title="fig : " ] ; at ( 7.3,.2 ) ; at ( 0,0 ) plotted as a function of @xmath20 for @xmath137 and @xmath138 ( blue , yellow , green , red and purple ) in the hypothetical case @xmath139 , @xmath140 for illustrative purposes .
the figure to the left shows the general behavior over the full range of @xmath20 , while that on the right shows the behavior in the dilute limit plotted in logarithmic scale where the long - range dependence on @xmath125 becomes relevant.,title="fig : " ] ; at ( 7.0,.2 ) ; our approach to obtain the above theorem is to first formulate the effects of the short- and long - range interactions in the form of local exclusion principles , an approach that goes back to dyson and lenard s original proof of the stability of matter for fermionic coulomb systems @xcite ( see also @xcite ) and which was developed in @xcite to prove lieb thirring inequalities ( cf .
@xcite ) with the usual pauli exclusion principle for fermions replaced by more general repulsive interactions for bosons .
the reason for the factor @xmath141 in compared to the expected value @xmath142 ( at least if comparing to fermions @xmath143 and assuming a linear interpolation to small @xmath26 s.t .
@xmath144 ) is that our long - range exclusion principle , which is applied locally on boxes of a tunable size , only increases linearly with the number of particles and is strongest on a scale where about two particles fit in each box .
we provide further bounds for @xmath129 in various parameter regimes in theorem [ thm : gas ] .
it should be remarked that our local exclusion principles also can be used to prove lieb thirring inequalities .
we postpone the extended case to future work but note that the ideal case is directly improved by the present results , namely replacing ( * ? ? ?
* lemma 8) with the local exclusion principle of lemma [ lem : localexclusionlr ] below yields the following bounds for ideal anyons , where the constant @xmath145 improves the one in ( * ? ? ?
* theorems 1 and 11 ) : [ thm : intro - ideallt ] + with @xmath9 , @xmath89 , @xmath84 and @xmath91 we have that @xmath146 and if @xmath147 is an external one - body potential , acting by @xmath148 , then @xmath149 for some positive universal constants @xmath150 and @xmath151 , and @xmath152 . the question concerning optimality of the above bounds with respect to their dependence on @xmath26 in the dilute limit is a very difficult one , and will be discussed elsewhere @xcite
. however , we would like to point out that it was suggested in @xcite ( see also @xcite ) that a class of fqhe - inspired trial states with a clustering behavior could minimize the energy for certain fractions , and here we find additional support for this claim ; cp . figure [ fig : rho / r_plot_clusters ] below .
furthermore , there was in @xcite , then based on the weaker bounds of @xcite , a slight inconsistency in the behavior with respect to odd - numerator @xmath26 which is remedied by the improved bounds presented here .
the structure of the paper is as follows .
we lay the foundations in sections [ sec : short - range ] and [ sec : relative ] by proving the short - range bound of lemma [ lem : short - range ] , and the basis for the long - range bound in the form of a relative magnetic hardy inequality with symmetry .
then the main body of the paper , section [ sec : long - range ] , is concerned with the application of this hardy inequality to prove the long - range bound of theorem [ thm : long - range ] .
this turns out to become surprisingly challenging in the extended case due to the oscillatory nature of the effective potential . in section [ sec : exclusion ] the long- and short - range bounds
are applied to prove local exclusion principles for anyons , and finally in section [ sec : gas ] we discuss the homogeneous anyon gas in the thermodynamic limit .
* acknowledgments . *
the authors would like to thank michele correggi , ari laptev , fabian portmann , nicolas rougerie , jan philip solovej and aron wennman for discussions .
financial support from the swedish research council , grant nos .
2012 - 3864 ( s. l. ) and 2013 - 4734 ( d. l. ) , is gratefully acknowledged .
the short - range interaction given by lemma [ lem : short - range ] comes as a simple consequence of the well - known magnetic inequality ( see e.g. ( * ? ? ?
* lemma 1.4.1 ) or @xcite ) @xmath153 splitting the coordinates according to @xmath154 for each particle @xmath50 , we write for the left - hand side of @xmath155 where we used the expression for @xmath156 in .
we have thus obtained the right - hand side of .
the inequality also follows directly from integrating the straightforward identity @xmath157 \pm { \mathbf{a}}\cdot { \nabla^{\perp}}|u|^2,\ ] ] with @xmath158 : = \frac{i}{2 } ( u\nabla\bar{u } - \bar{u}\nabla u ) $ ] .
we note that the dirichlet boundary conditions on @xmath159 are in fact necessary here since the bound is otherwise invalid , as can be seen by taking @xmath160 , @xmath161 , and the trial state @xmath162 .
similarly , had we considered the inequality locally on a small enough domain ( compared to @xmath17 ) we would have also found a contradiction as @xmath31 , unless dirichlet boundary conditions are enforced .
for the long - range statistical interaction between anyons we take the same starting point as in @xcite , namely , the core observation is the validity of a relative magnetic hardy inequality which respects the symmetry of the anyon problem .
non - symmetric versions of this inequality were introduced and studied in @xcite ( one - particle version ) and in ( * ? ? ? * theorem 2.7 ) ( many - particle version ) ; see also @xcite , ( * ? ? ?
* chapter 5.5 ) and references therein . however , as was pointed out in @xcite , symmetry is crucial in order to obtain non - trivial bounds in the many - particle limit .
we formulate the following version of the inequality quite generally . initially , consider a magnetic field @xmath163 defined on a disk of radius @xmath118 , and assumed to be determined by a suitable continuous vector potential @xmath164 as @xmath165 .
then the normalized flux inside a smaller disk of radius @xmath166 is given by @xmath167 note that if we were only given @xmath168 on some annulus @xmath169 , with @xmath170 , i.e. if we only knew @xmath171 on @xmath102 ( so that only the right - hand side of makes sense for @xmath172 ) , then @xmath171 can nevertheless be extended ( non - uniquely ) to the full interior @xmath173 , for example by taking @xmath174 with @xmath175 here defined in terms of @xmath176 as in ( note that we are not considering extending @xmath176 ) . then both expressions for @xmath177 in are well defined and agree for all @xmath172 .
we also note that if the magnetic field is antipodal - symmetric on @xmath102 , i.e. @xmath178 for all @xmath179 , then the corresponding potential must ( if gauge - normalized correctly ) be antipodal-__anti__symmetric , @xmath180 , @xmath179 , and vice versa .
[ lem : magnetic - hardy - symm ] let @xmath181 , with @xmath182 , be an annular domain in @xmath73 .
let @xmath168 be a continuous vector potential corresponding to a magnetic field @xmath171 , @xmath183 , that is defined on the entire disk @xmath184 such that the normalized flux @xmath177 given by is finite for all @xmath185 .
furthermore , assume that @xmath176 is antipodal-*antisymmetric * resp . @xmath171 is antipodal-*symmetric * on @xmath102 , i.e. @xmath180 resp .
@xmath178 for @xmath179 .
then , for any antipodal-*symmetric * @xmath186 , i.e. with @xmath187 for all @xmath179 , @xmath188 alternatively , if instead @xmath159 is antipodal-*antisymmetric * , @xmath189 for all @xmath179 , then @xmath190 we apply the techniques from @xcite , with symmetry taken into account as in ( * ? ? ?
* lemma 2 ) .
we start by letting @xmath191 $ ] denote the magnetic quadratic form on @xmath102 , @xmath192(u ) : = \int_\omega |(-i\nabla+{\mathbf{a}})u|^2\ , d{\mathbf{r}}= \int_{r_1}^{r_2}\!\!\int_0^{2\pi } \bigl(|(-i\partial_r + a_r)u|^2 + r^{-2}|(-i \partial_{\varphi}+ r a_{\varphi } ) u|^2\bigr ) r\ , d{\varphi}dr,\ ] ] where @xmath193 and @xmath194 . for the first term above we use the diamagnetic inequality @xmath195 , while for the second we can for each @xmath185 explicitly diagonalize the self - adjoint operator @xmath196 acting on @xmath197 .
the corresponding eigenvalues and normalized eigenfunctions of this operator are given by : @xmath198 for @xmath199 .
because of the antipodal - antisymmetry of @xmath176 , implying antipodal - symmetry of @xmath200 , i.e. @xmath201 , we have that @xmath202 therefore , only the even / odd terms will contribute upon expanding @xmath203 as @xmath204 with @xmath205 and @xmath206 . by the above remarks and parseval s identity
we find that @xmath207(u ) & = \int_{r_1}^{r_2}\!\ !
\int_0^{2\pi } |(\partial_r + ia_r)u|^2 \ ,
r\ , d{\varphi}dr + \int_{r_1}^{r_2 } \sum_{k\in \z_{e / o } } |\lambda_k(r)|^2 |u_k(r)|^2 \ , r^{-1}\ , dr \\ & \geq \int_{r_1}^{r_2}\!\
! \int_0^{2\pi }
|u|\bigr|^2 \ , r\ , d{\varphi}dr + \int_{r_1}^{r_2 } \inf_{k\in \z_{e / o } } |\lambda_k(r)|^2 \sum_{k\in \z_{e / o } } |u_k(r)|^2 \ , r^{-1}\ , dr \\ & = \int_{r_1}^{r_2}\!\ ! \int_0^{2\pi } \bigl(\bigl|\partial_r |u|\bigr|^2 + r^{-2}\inf_{k\in \z_{e / o } } |\lambda_k(r)|^2 |u|^2\bigr ) r\ , d{\varphi}dr .
\end{aligned}\ ] ] thus the estimate we are left with is @xmath208(u ) \geq \int_{r_1}^{r_2}\!\ ! \int_0^{2\pi }
\bigl(\bigl|\partial_r |u|\bigr|^2 + r^{-2}\inf_{k\in \z_{e / o}}\bigl|\hat\phi(r)-k\bigr|^2 |u|^2\bigr ) r\ , d{\varphi}dr,\ ] ] which proves the lemma .
the above lemma not only extends the inequality of ( * ? ? ?
* lemma 2 ) to more general ( extended ) magnetic fields , but also improves it by keeping the radial derivative .
this turns out to be crucial in order to obtain an improved dependence on @xmath26 in the dilute limit .
we note that also @xcite discarded the radial derivative in their inequalities .
we set out to prove theorem [ thm : long - range ] , and first note that by the remarks in section [ sec : intro - prelims ] we may assume without loss of generality that @xmath209 .
proceeding as was done in @xcite for the non - extended case @xmath89 , we start from the expression for the kinetic energy on a domain @xmath90 , @xmath210 and we are considering the first @xmath211 particles @xmath212 while the remaining @xmath213 ones may reside anywhere in @xmath73 . using that , for any @xmath214 ,
@xmath215 we have that @xmath216 where we also note that the magnetic field present in the last ( total momentum ) term simplifies to @xmath217 by the antisymmetry of the vector potential , and thus vanishes if @xmath218 .
we now study the inner integral in for the @xmath219 particle pair , and introduce relative coordinates : @xmath220 giving @xmath221 where the relative vector potentials are given by @xmath222 hence , for any positions @xmath223 of the other particles and for each center - of - mass coordinate @xmath224 of the particle pair , we observe that the resulting magnetic vector potential @xmath225 is antipodal - antisymmetric on the relative disk @xmath226 with a corresponding antipodal - symmetric magnetic field @xmath227 ( given for @xmath118 ) . also , the smooth function defined relative to @xmath228 by @xmath229 is antipodal - symmetric on @xmath230 .
hence , we may apply the relative magnetic hardy inequality of lemma [ lem : magnetic - hardy - symm ] ( for @xmath89 we split into concentric annuli avoiding the @xmath231 as in ( * ? ? ? * theorem 4 ) ) to obtain that @xmath232 where @xmath233 @xmath234 and we from now on express the @xmath235 other particles @xmath231 relative to the fixed center of mass @xmath228 of the considered pair : @xmath236 .
the problem has thus boiled down to study this explicit radial _ scalar _
interaction potential @xmath237 , which is essentially an inverse - square repulsion modulated with a coupling strength @xmath238 which measures how well the normalized flux @xmath177 stabilizes away from the even integers . in the dilute situation the flux and hence also @xmath239 would for the most part be constant , however we could have significant oscillations of @xmath238 between one and zero whenever many particles are enclosed over short differences in the radial variable @xmath240 ( see figure [ fig : rho / r_plot ] ) .
controlling these oscillations turns out to be a significant challenge , and the entire remainder of this section shall be concerned with proving the following theorem , from which theorem [ thm : long - range ] follows .
[ thm : mainradialbound ] for any @xmath241 , @xmath242 $ ] , @xmath243 , @xmath244 , rdr)$ ] , and with @xmath239 defined as above , we have that @xmath245}|u|^2 \biggr ) r\ , dr,\ ] ] with @xmath114 .
in the case @xmath89 we may take @xmath111 .
[ rem : mainradialbound ] the margins which appear here as a cut - off for the potential are not optimal and could be improved with more work , to the cost of an even weaker constant .
the main reason for the weakness of the constant @xmath246 is the fact that we have chosen to control the above form by means of filling the gaps around the zeros of the potential by smearing it over longer ( but not too long ) intervals , and that in the worst possible situation there are very large regions of intense oscillation and many such zeros . by considering the special case @xmath247 and
densely packed , overlapping particles ( i.e. when @xmath248 is large ) distributed so that the effective magnetic field is approximately constant , we find that @xmath246 can not be greater than @xmath249 , which is what the corresponding constant would be if one applied the same argument to the case of a homogeneous magnetic field ( see below ) . however , for @xmath250 ( or small enough so that @xmath239 is larger than @xmath251 for a sufficiently large set of radii ) , we expect that the ground - state energy of the left - hand side ( though difficult to compute in general ) should in almost all situations be bounded by that with @xmath238 replaced by @xmath252 ( compare figure [ fig : rho / r_plot ] ) .
we discuss further possible improvements to the constant @xmath246 at the end of section [ sec : prooflongrange ] . at ( 0,0 ) , now for @xmath253 particles in a disk of radius @xmath254 with @xmath255 ( @xmath40 ) , and with a single particle close to our center of mass and the remaining nine in clusters of three .
note that in this case the effective potential can become identically zero on long intervals.,title="fig : " ] ; at ( 5.65,0 ) ; at ( 0.05,3.6 ) ; at ( 0,0 ) , now for @xmath253 particles in a disk of radius @xmath254 with @xmath255 ( @xmath40 ) , and with a single particle close to our center of mass and the remaining nine in clusters of three .
note that in this case the effective potential can become identically zero on long intervals.,title="fig : " ] ; at ( 5.65,0 ) ; at ( 0,3.6 ) ; at ( 0,0 ) , now for @xmath253 particles in a disk of radius @xmath254 with @xmath255 ( @xmath40 ) , and with a single particle close to our center of mass and the remaining nine in clusters of three .
note that in this case the effective potential can become identically zero on long intervals.,title="fig : " ] ; at ( 5.65,0 ) ; at ( 0.05,3.6 ) ; at ( 0,0 ) , now for @xmath253 particles in a disk of radius @xmath254 with @xmath255 ( @xmath40 ) , and with a single particle close to our center of mass and the remaining nine in clusters of three .
note that in this case the effective potential can become identically zero on long intervals.,title="fig : " ] ; at ( 5.65,0 ) ; at ( 0,3.6 ) ; inserting the bound of theorem [ thm : mainradialbound ] with @xmath256 into the expressions , , and , we obtain the first bound of the theorem .
furthermore , by rescaling @xmath257 and considering the minimizer @xmath258 which is the solution of the bessel equation @xmath259 with the minimal eigenvalue @xmath260 , one obtains @xmath261 and therefore , after the simplifying estimate @xmath262 , the second bound of theorem [ thm : long - range ] .
the properties of @xmath115 described in the theorem are direct consequences of proposition [ prop : besseleigenvaluebounds ] . before continuing with the proof of theorem [ thm : mainradialbound ] we note that , although this method involving the magnetic hardy inequality turns out to be sufficient and indeed well - suited for our purposes , it does not deal well with strong magnetic fields ( hence also the presence of a large external field ) , as the following example shows .
the strong magnetic fields arising from a large overlap between the particles will instead be handled by the short - range part of the interaction , lemma [ lem : short - range ] .
[ prop : constant - field ] the ground - state energy @xmath263 for the neumann form ( with no symmetry imposed ) with a constant magnetic field @xmath264 on the unit disk , @xmath265 satisfies @xmath266 however , the ground - state energy for the corresponding lower bound obtained from the hardy inequality , @xmath267 is bounded from above by @xmath268 independent of @xmath269 .
the first estimate follows e.g. from ( * ? ? ?
* theorem 5.3.1 ) , while the second from bounding the infimum by @xmath141 and taking as a trial state the bessel function @xmath270 .
our strategy in order to find a uniform bound for the scalar interaction of theorem [ thm : mainradialbound ] will be to borrow a bit of the radial kinetic energy to smear @xmath239 over intervals whenever it has critical oscillations . as a preliminary to the proceeding analysis we therefore study the localized effective quadratic form @xmath271,\ ] ] on an interval @xmath272 , and our goal is to find a bound of the form @xmath273 i.e. corresponding to @xmath239 being constant .
[ lem:1dprojection ] let @xmath274 be an interval @xmath275 , such that @xmath276 and @xmath277 , and let @xmath278 be non - negative with @xmath279 .
then for any @xmath242 $ ] we have that @xmath280 where @xmath281 denotes the weighted mean on @xmath274 , @xmath282 and @xmath283 is an explicit function satisfying @xmath284 .
this lemma can be proven under more general conditions ; the only condition on @xmath274 needed for our proof is that @xmath285 is sufficiently small .
the current setting is simply what we require later . by the change of variables @xmath286 and with @xmath287
we find that @xmath288 for this quadratic form we can perform a projection - type argument to bound the first eigenvalue of the associated operator @xmath289 ( with neumann boundary conditions ) , which in turn will imply a bound of the desired form .
let @xmath290 denote the orthogonal projection onto the ground state @xmath291 of @xmath292 , where @xmath293 , and let @xmath294 .
then @xmath295 and @xmath296 ( the first non - zero neumann eigenvalue of @xmath292 ) .
since @xmath297 , an application of cauchy
schwarz and young s inequalities yields , for any @xmath298 and @xmath299 , that @xmath300 hence we see that @xmath301 the first of these operators is equal to @xmath302 , where @xmath303 and the second we bound from above by @xmath304 .
we find that for any @xmath299 the operator @xmath305 satisfies @xmath306 with @xmath307 and @xmath276 we find that @xmath308 hence , writing @xmath309 , @xmath310 , and using that @xmath311 , @xmath312 we have that @xmath313 where we assumed the positivity of the second argument ( this will be clear by the choice of @xmath269 below ) . note that the first argument of the minimum is decreasing in @xmath314 while the second one is increasing .
thus to find the maximizing @xmath269 we only need to solve the equation @xmath315 . plugging the solution , given by @xmath316 into the above yields @xmath317
finally , since @xmath283 is a convex function for @xmath242 $ ] we can simplify this expression using @xmath318 and by simple numerical estimates one finds that @xmath319 . to proceed with the analysis we will need a more precise understanding of how @xmath239 depends on the positions of the other particles .
note first that we may assume that @xmath320 using the reflection - conjugation symmetry .
we then begin by writing for the normalized flux @xmath321 where we introduce the _ particle counting function _ @xmath322 and , as already mentioned , all particles have been renumbered and expressed relative to the fixed center of mass @xmath228 of the considered particle pair : @xmath323 .
in terms of this function we have that @xmath324 and we may cover the interval @xmath325 $ ] by smaller intervals @xmath326 labeled by the minimizer @xmath327 ( note the monotonicity of the function @xmath328 , and that we might already have @xmath329 on the first such interval at @xmath330 ) .
each @xmath326 contains , except possibly for the first and last such interval , exactly one zero of @xmath239 which we denote by @xmath331 : @xmath332 so that @xmath333 we then also have the very useful identity @xmath334 whenever @xmath335 .
let us denote by @xmath336 and @xmath337 the nearest points to the left resp .
right of @xmath331 where @xmath338 and @xmath339 $ ] unless @xmath239 stabilizes at @xmath340 on some interval between @xmath331 and @xmath341 , in which case @xmath342 and the intervals @xmath326 and @xmath343 overlap . ] , then @xmath344 finally , we also denote by @xmath345 and @xmath346 the nearest points to the left resp .
right of @xmath331 where @xmath347 , and hence @xmath348 , and we observe due to and that @xmath349 recall that this constant depends in a non - trivial way on number - theoretic aspects of the parameter @xmath26 , and that it remains bounded away from zero for all @xmath0 if and only if @xmath26 is an odd - numerator rational number ( see ( * ? ? ?
* proposition 5 ) ) .
hence , we can reduce our problem to studying precisely those smaller intervals around each zero of @xmath239 not covered by .
denote by @xmath350 , with @xmath351 , any such interval around a zero @xmath352 $ ] on which @xmath353 and observe by the size of each particle that @xmath354 , and furthermore that there is always at least one particle covering the entire interval : [ lem : oneparticlecoverik ] if @xmath355 is a zero of @xmath239 then with @xmath356 constructed as above there exists a particle centered at @xmath357 , at a distance @xmath358 , such that @xmath359 $ ] .
in other words , the angular projection of some particle completely covers @xmath356 .
let @xmath360 and let @xmath361 be the particle counting function corresponding to our particle configuration but where we remove all particles ( seen as closed disks @xmath362 ) that have empty intersection with the closed disk @xmath363 , i.e. we remove all particles that are centered at a distance strictly larger than @xmath364 from the origin . by the construction of @xmath356
, there is at least one particle that has non - empty intersection with @xmath365 ( not counting any fully enclosed ones ) , since otherwise @xmath328 would be constant here which contradicts the choice of @xmath366 .
let now @xmath367 be the radius such that all the particles that intersected @xmath365 are completely contained in the closed disk @xmath368 . by the construction of @xmath361 ,
its value at @xmath367 is an integer which , since there were particles intersecting @xmath365 , is strictly larger than @xmath369 .
but then , since @xmath370 , the function @xmath328 must take at least one integer value on @xmath371 $ ] .
thus , by the definition of @xmath372 we conclude that @xmath373 , which completes the proof .
to proceed we will need more information on the local behavior of the particle counting function @xmath328 .
we note that @xmath374 where @xmath375 are the centers ( in relative coordinates ) of the @xmath235 particles not in our presently studied pair . to analyze @xmath328 we thus need to work with the area of the intersection of pairs of disks .
an elementary , although slightly tedious , calculation yields the following expression .
[ prop : the_area_of_intersecting_disks ] let @xmath376 and @xmath377 be disks of radii @xmath378 , with @xmath379 , centered at the points @xmath380 and @xmath381 .
then with @xmath382 we have for the area of intersection , in the non - trivial regime @xmath383 and @xmath384 , that @xmath385 if @xmath386 the area is zero and if @xmath387 the area is @xmath388 . differentiating the flux contribution from a single particle located at @xmath389 , given by @xmath390 we find for arbitrary @xmath391 that @xmath392 0 , & \textrm{if } r
> r+d \textrm { or } r < d - r , \\ \frac{2 r}{\pi r^2}\arccos\bigl(\frac{d^2+r^2-r^2}{2 d r}\bigr ) , & \textrm{otherwise . } \end{array}\right.\ ] ] we first note that this function is concave on its support ( compare figure [ fig : particleprofile ] ) , as shown explicitly in appendix [ app : concavity ] .
furthermore , it satisfies some simple bounds : [ lem : profilebounds ] with @xmath393 denoting the one - particle profile we have for @xmath394 and @xmath395 the following bounds : @xmath396 the upper bound for @xmath397 given by the lemma is clear from the geometric construction of @xmath397 and @xmath398 .
the value of @xmath397 is equal to the length of the circle segment @xmath399 where @xmath400 , divided by @xmath401 , and clearly this can not exceed @xmath402 . for the lower bound we use concavity .
for @xmath403 the function @xmath393 is concave on its support @xmath404 $ ] ( see appendix [ app : concavity ] ) . since @xmath405 has the same support and it is continuous , piecewise linear and concave on this interval with maximum at @xmath406 it suffices to prove that @xmath407 : for @xmath403 we have that @xmath408 is a decreasing function and that @xmath409 .
for @xmath410 we have that @xmath393 and @xmath411 are concave on @xmath412 $ ] and zero otherwise ( see appendix [ app : concavity ] ) . by the linearity of @xmath413 on this interval it is sufficient to prove that @xmath414 , which follows since @xmath415 . at ( 0,0 ) and its lower bound @xmath411 plotted over the support of @xmath397 .
the profile depicted is for @xmath416 , while as @xmath417 increases this profile more and more resembles the upper half of a disk.,title="fig : " ] ; at ( 0,0.05 ) ; at ( 3.85,0.05 ) ; at ( 4.5,0 ) ; at ( 7.2,0.05 ) ; the following lemma captures in a convenient form essential aspects of the shape of the particles , and will play an important role in the analysis on intervals of oscillation below .
[ lem : shapelemma ] if @xmath418 $ ] with @xmath276 and @xmath419 , we have that @xmath420 the assumption @xmath421 can be relaxed slightly by instead requiring that @xmath422 is sufficiently large . in particular , in the limit @xmath423 the one - particle profile approaches a half disk and it is then geometrically clear that the statement holds whenever @xmath424 .
it is sufficient to prove that this holds for the one - particle profile @xmath393 for any @xmath394 .
the proof utilizes that the profile @xmath393 is concave on its support intersected with @xmath425 , which is shown in appendix [ app : concavity ] . if in addition @xmath403 the profile is concave on its full support @xmath426 ,
also shown in appendix [ app : concavity ] .
thus , whenever @xmath275 does not contain the maximum of @xmath393 the statement is clear , since if this is the case @xmath393 is monotone here and thus has its maximum value in either @xmath422 or @xmath427 .
thus we may assume that the unique maximum of @xmath393 is attained at a point @xmath428 in @xmath275 .
moreover , by the concavity of @xmath393 it suffices to consider the case when @xmath429 .
the inequality we wish to prove can now be written as @xmath430 which should hold for all @xmath276 such that @xmath431 .
* case 1 : @xmath403 . * in this case it holds that @xmath432 , since @xmath433 .
this can be verified by considering @xmath434 , which can be shown by straightforward computation to be decreasing in @xmath417 and moreover it is negative at @xmath435 .
similarly , @xmath436 can be verified to be positive , and hence @xmath437 .
this implies that the right - hand side of is a concave function of @xmath422 , and hence its minimum value is attained at one of the extremal points of the allowed intervals .
but this is precisely when either @xmath422 or @xmath427 is equal to @xmath428 , in which case the statement is trivial by the non - negativity of @xmath397 .
* case 2 : @xmath438 .
* by similar calculations as in case 1 , we have that @xmath439 for @xmath440 . then by concavity @xmath441 is a monotonically decreasing function on @xmath412 $ ] .
thus @xmath442 and the statement follows .
* case 3 : @xmath443 . *
again the function @xmath393 is concave on @xmath444 .
thus we again only need to consider the extremal cases of the intervals @xmath275 containing the maximum of @xmath393 on this interval .
this reduces to three different options .
either @xmath445 , or @xmath446 , or one of the endpoints of the interval is located at the maximum . in the last case
the statement is trivially true .
if we were in the second option then @xmath447 .
through a similar computation as above one checks that on this interval @xmath393 is monotone , and hence the statement follows ( it turns out that this interval never actually contains the maximum and could thus also be excluded for this reason ) .
if however @xmath448 the inequality is reduced to @xmath449 by scaling we may without loss of generality assume that @xmath450 . using the explicit expression of
@xmath397 we need to show that @xmath451 since for @xmath452 we have that @xmath453 , it follows that @xmath454 but the function @xmath455 is decreasing in @xmath240 , for @xmath456 , and thus we only need to verify the inequality @xmath457 this is equivalent to @xmath458 .
we observe that the left - hand side of this inequality is decreasing whilst the right is increasing .
thus it suffices to check the validity at @xmath459 , which is a simple numerical evaluation . in this subsection
we use the explicit form of @xmath328 uncovered above for @xmath460 and the projection argument of lemma [ lem:1dprojection ] to locally replace the effective 1-dimensional potential @xmath461 with some constant times @xmath462 . by lemma [ lem:1dprojection ]
it suffices to prove that given an interval @xmath463 of small enough measure we have a suitable bound for the weighted mean @xmath281 on @xmath274 . on intervals where @xmath239 is already larger than @xmath252 we need not perform any detailed analysis .
thus the only intervals that remain are those of the form @xmath464 close to the zeros of @xmath239 .
the analysis is split into several parts depending on the behavior of @xmath239 near a specific zero .
our first bound provides a general estimate for @xmath281 on any subinterval of the @xmath326 constructed above ( section [ sec : potential - structure ] ) which contains the unique zero of @xmath239 on this interval .
[ lem : generalintegralbound ] let @xmath275 , with @xmath465 , be such that on this interval @xmath466 for some fixed @xmath467 and such that there exists some @xmath468 with @xmath469 . then , with @xmath470 , we have that @xmath471 where as before @xmath328 denotes the particle counting function . on such an interval @xmath275
we can , according to , express @xmath239 in terms of @xmath472 as @xmath473 inserting this into the integral we wish to bound and using the trivial estimate @xmath474 , we have that @xmath475 we split the above integral into two parts , @xmath476 using the cauchy
schwarz inequality and changing the order of integration one finds that @xmath477 to obtain the desired estimate we combine the above with the observation that both @xmath478 and @xmath479 are larger than @xmath480 , and the elementary inequality @xmath481 , @xmath482 we now study @xmath281 on the intervals @xmath483 constructed earlier around zeros of @xmath239 , with @xmath484 .
we begin with a lemma providing a bound for the local weighted mean on a certain subclass of these intervals where the potential is in some sense well behaved .
[ lem : goodintervals ] let @xmath360 be one of the intervals constructed above which satisfies @xmath485 . then if either @xmath486 for some @xmath487 , we have that @xmath488 we will later see that for our treatment of intervals @xmath356 that are not covered by this lemma we will need to choose @xmath150 rather small , approximately @xmath489 . by lemma [ lem : generalintegralbound ]
we may estimate the integral of the potential by @xmath490 by lemma [ lem : oneparticlecoverik ] the interval @xmath356 is covered by at least one particle .
thus for @xmath491 we can bound @xmath492 from below by using our lower bound for the one - particle profile @xmath493 and minimizing over particle positions @xmath417 such that @xmath494 .
let as before @xmath495 denote the lower bound for @xmath397 given by lemma [ lem : profilebounds ] .
we conclude that @xmath496 as this integrand is piecewise linear in @xmath417 we must have that the integral is minimized in one of the extremal points : a particle starting at @xmath366 , a particle ending at @xmath372 or a particle centered at @xmath497 . by symmetry
the last alternative maximizes the integral and thus we can discard this option .
moreover , the same symmetry implies that the first two alternatives are equal . through a straightforward calculation
we find that @xmath498 |i_k| , & \textrm{if } |i_k| > r. \end{array}\right.\ ] ] thus if @xmath499 , @xmath500 , the above yields @xmath501 if instead of @xmath499 we have that @xmath502 we can obtain the same bound .
namely , if we again consider the bound given by lemma [ lem : generalintegralbound ] , @xmath503 we find , using @xmath504 , that @xmath505 where we also used that @xmath506 for each @xmath507 . for the weighted mean we now find that @xmath508 where we used that @xmath509 and @xmath485 implies that @xmath510
. the previous lemma does not cover the scenario where @xmath328 increases rapidly , resulting in rapid oscillations on many short intervals @xmath356 . in the next lemma we consider the remaining intervals @xmath356 and use our geometric knowledge of @xmath328 to show that these intervals can not cover too much of our large - scale interval @xmath511 $ ] . to achieve this we first cover the remaining collection of intervals @xmath356 with a collection of intervals @xmath512 such that @xmath513 for all @xmath514 .
[ lem : badikhassmallmeasure ] let @xmath515 $ ] be an interval of length @xmath516
. then the fraction of @xmath517 covered by intervals @xmath356 satisfying both @xmath518 with @xmath519 , is less than @xmath520 let @xmath521 denote the set of intervals such that is satisfied , @xmath522 for each @xmath523 , and ordered from left to right . for notational convenience
we will let @xmath524 and @xmath525 denote @xmath526 and @xmath527 , respectively .
we will also denote by @xmath528 and @xmath529 a ( fixed ) choice of points in each @xmath356 such that @xmath530 and @xmath531 .
we begin by showing that we may assume that the distance between any two points in two consecutive intervals is less than @xmath516 , allowing us to apply lemma [ lem : shapelemma ] . if , for some @xmath532 , @xmath533 and @xmath534 are such that @xmath535 , then since both intervals have non - empty intersection with @xmath517 we must have that @xmath536 . but
this implies that @xmath537 and the statement follows .
similarly the statement is true if @xmath538 .
suppose that there exists a @xmath50 such that @xmath539 .
then , since we by the above may assume that @xmath540 , lemma [ lem : shapelemma ] implies that @xmath541 but combined with this leads to a contradiction : @xmath542 which is impossible since @xmath543 . let us say that an interval @xmath356 where @xmath544 is of type a , and one where instead @xmath545 is of type b ( note that @xmath546 by the assumption on @xmath356 ) .
we let @xmath547 and @xmath548 denote the collections of intervals of type a and type b respectively . the above contradiction argument yields that an interval of type a can not follow one of type b , i.e. if we for some @xmath50 have that @xmath549 then @xmath550 for all @xmath551 , and similarly , if @xmath552 then @xmath553 for all @xmath554 .
we conclude that there is at most one @xmath507 such that @xmath356 and @xmath555 are of different type , and @xmath356 must then be of type a. as we will now show , it turns out that the sequence of lengths @xmath556 of consecutive intervals starting at any interval of type a and going to the left , resp .
type b and going to the right , is monotonically decreasing and bounded from above by a geometric sequence . by assumption , all @xmath557 , and in particular
this holds for the first interval in any such sequence . using these observations
we will be able to bound the total measure of @xmath558 .
we begin by studying a sequence starting at an interval of type a and going to the left ( note that such a sequence may not exist if all @xmath559 ) .
we wish to prove that @xmath556 decreases along this sequence .
let @xmath50 be such that @xmath560 .
then @xmath561 , and by lemma [ lem : shapelemma ] we have that @xmath562 .
since we assume that @xmath563 this implies that @xmath564 the only thing we used above was that @xmath565 . since this implies that also @xmath566 , we can iterate this argument until we reach @xmath567 . this yields for @xmath551 that @xmath568 using that @xmath569 ( for any @xmath507 ) we , for @xmath551 , find that implies @xmath570 where we used that , for @xmath571 , @xmath572 since otherwise @xmath573 would be zero which can not happen by the construction of the @xmath356 s . since @xmath150 is small
this proves the claim in the case of type a intervals . for the case of type b
intervals the proof is almost identical and one finds instead that , if @xmath574 , @xmath575 we are now ready to complete the proof of the lemma .
begin by finding @xmath50 such that @xmath560 and @xmath576 ( if @xmath547 , alt .
@xmath548 , is the empty set we take @xmath577 , alt .
@xmath578 ) . then using the above estimates we obtain that @xmath579 and dividing this quantity by @xmath580 completes the proof .
what we have found is that the measure where @xmath239 is already large or can be averaged to be large in @xmath517 is at least @xmath581 using this we can find a non - trivial uniform lower bound on @xmath582 and therefore , using the local projection argument , we finally obtain that there exists a constant @xmath583 such that @xmath584}|u|^2\bigr)r\ , dr.\ ] ] we proceed as follows : @xmath585 where @xmath586 denotes a new weight obtained by replacing @xmath238 with @xmath587 on all @xmath356 covered by lemma [ lem : goodintervals ] that intersect @xmath588 , by using lemma [ lem:1dprojection ] with @xmath589 .
thus the only remaining zeros of @xmath586 on @xmath588 are those contained in intervals @xmath356 which satisfy the assumptions of lemma [ lem : badikhassmallmeasure ] .
we can now cover all remaining @xmath356 in @xmath590 by a collection of disjoint intervals @xmath517 of length @xmath580 and which are contained in @xmath588 .
for instance we may take the intervals @xmath591 where @xmath211 runs from @xmath340 to @xmath592 . on each such @xmath593 we then have that @xmath594 by lemma [ lem : badikhassmallmeasure ] we then obtain for the weighted mean of @xmath586 that @xmath595 thus for each @xmath517 we can again apply lemma [ lem:1dprojection ] and obtain @xmath596 applying this for each @xmath517 we obtain the desired estimate with @xmath597 maximizing this in @xmath598 we obtain for @xmath599 the extremely small ( but positive ) constant @xmath600 this concludes the proof of theorem [ thm : mainradialbound ] and hence the treatment of the long - range interaction of theorem [ thm : long - range ] .
we note that with this choice of @xmath150 we allow for approximately @xmath601 of any ( and all ) @xmath516 long interval contained in @xmath602 $ ] to be covered by the intervals @xmath356 satisfying . as we expect that this is rather far from the actual situation for most particle configurations there seems to be room for improvement in the above considerations .
one such improvement could be to use that the effective potential must between every two @xmath356 intervals go up to one and then back down again
. our current method does not take this into account and is blind to the fact that there must be helpful gaps between the @xmath356 s .
another way of improving this constant would be to refine the bounds in lemma [ lem : goodintervals ] by using the precise shape of the one - particle profile instead of the simpler lower bound provided by @xmath405 .
one could also take into account that all intervals can not be at the edge of a particle , i.e. make use of the observation that a large number of the particles are likely to cover more than one interval @xmath356 .
we now formulate the obtained energy bounds for anyons in terms of local exclusion principles , following @xcite . with a weight partition @xmath603 ^ 3 $ ] , @xmath604 , we can write for the total kinetic energy for @xmath0 anyons in a normalized state @xmath96 @xmath605 where we used lemma [ lem : diamagnetic ] and lemma [ lem : short - range ] .
we then make a partitioning of the plane @xmath73 into disjoint squares @xmath606 s : @xmath607,\ ] ] where the expected local energy on each square @xmath606 is given by ( the definitions extend to all @xmath608 ) @xmath609 & : = \sum_{j=1}^n \int_{\r^{2n } } \biggl ( \kappa_1 \bigl|\nabla_j|\psi|\bigr|^2 + \kappa_2 \sum_{\substack{k=1\\k \neq j}}^n 2\pi|\alpha| \frac{{{\ensuremath { \mathds 1 } } } _ { b_r(0)}}{\pi r^2}({\mathbf{x}}_j - { \mathbf{x}}_k ) \ , |\psi|^2 + \kappa_3 |d_j\psi|^2 \biggr ) { { \ensuremath { \mathds 1 } } } _ q({\mathbf{x}}_j ) \ , d{\textup{x}}\\[-20pt ] & \phantom{:}\ge \sum_{n=0}^n e^{{\boldsymbol{\kappa}}}_n(|q| ) p_n(\psi;q).\end{aligned}\ ] ] here the local @xmath211-particle energy ( translation invariant and with neumann b.c . )
is given by @xmath610 and @xmath611 denotes the @xmath211-particle probability distribution induced from @xmath7 , @xmath612 having the normalizations @xmath613 and @xmath614 , the expected number of particles on @xmath606 . in the operators @xmath99 still depend on all @xmath0 particles , with the first @xmath211 on @xmath606 , and we take the infimum over the remaining @xmath213 positions in @xmath615 .
we consider first the contribution to the local energy coming solely from the short - range part of the magnetic interaction .
[ lem : localexclusionsr ] for any @xmath9 , @xmath59 and @xmath616 a square , and with @xmath617 , we have that @xmath618 and @xmath619 \ge \frac{{e_{\mathrm{sr}}}(\alpha , \gamma(q ) , \int_q\varrho_\psi)}{|q| } \biggl ( \int_q\varrho_\psi \ -1 \biggr)_{\!\!+},\ ] ] where @xmath620 2|\alpha| \gamma^{-2 } n & \text{for $ \gamma \ge \sqrt{2}$. } \end{array}\right.\ ] ] here @xmath621 and @xmath622 denotes the modified bessel function of order @xmath623 .
we consider the local energy form in . in the case
that @xmath624 , the short - range potential in the second term covers the full domain @xmath606 for every particle , and hence @xmath625 by convexity we then also have that @xmath626 in the case that @xmath627 , we use dyson s lemma @xcite in two dimensions ( see @xcite ) to smear the potential to the full domain as done in ( * ? ? ?
* proposition 19 ) , keeping part of the potential intact and smearing the rest . for @xmath628 and any @xmath242 $ ] we can bound the energy form in @xmath629 from below by @xmath630 with @xmath631}(r).\ ] ]
this expression arises from the application of dyson s lemma ( * ? ? ?
* lemma 3.1 ) on the star - shaped domain @xmath632 with the requirement that @xmath633 and where the considered pair potential is @xmath634 with scattering length ( see e.g. ( * ? ? ?
* appendix a.2.4 ) ) @xmath635 we now demand that @xmath636 be chosen such that the potentials match @xmath637 that is , @xmath638 however , note that the factor @xmath639 in @xmath70 diverges as @xmath640 while the other potential term stays bounded , implying @xmath641 . hence , in order to be able to bound @xmath642 uniformly we instead truncate the potential @xmath70 by replacing the unbounded factor with @xmath643,\ ] ] also using that @xmath135 ( see ( * ? ? ?
10.33.1 ) ) . with this replacement in the above
we then find that @xmath644 and hence @xmath645 and @xmath646 .
summing up , we find for all @xmath647 that @xmath648 and may again use convexity in @xmath211 to obtain the corresponding bound for @xmath649 $ ] .
although not aiming to provide the sharpest possible bound , the above lemma has the advantage of being relatively simple and it captures the overall dependence of the pure short - range interaction on the parameters . in a certain regime however , referred to below as the _ soft - core regime _ , the following version ( which could in some sense be viewed as a mix between the @xmath650 and @xmath138-dimensional cases studied in @xcite ) will yield a comparatively good bound . [
lem : templesmearbound ] for any @xmath16 and @xmath616 a square , and with @xmath617 , we have that @xmath651 for any @xmath652 , @xmath9 and @xmath653 such that @xmath654 .
following @xcite we write for the operator of the left - hand side @xmath655 with @xmath656 we apply the following result due to temple @xcite : if @xmath657 , for some schrdinger operator @xmath658 and scalar potential @xmath659 , then the ground - state energy of @xmath660 is bounded from below by @xmath661 as long as @xmath662 is positive .
here @xmath663 denotes the normalized ground state of @xmath664 , @xmath665 is the expectation of @xmath666 in the state @xmath663 , and @xmath667 resp .
@xmath668 is the first resp .
second eigenvalue of @xmath664 . in our case @xmath669
and we have that @xmath670 where for the lower bound one integrates the first particle of each pair on a smaller domain .
moreover , by cauchy
schwarz @xmath671 thus temple s inequality yields that @xmath672 as claimed .
we now turn to local energy bounds for the pure long - range part of the magnetic interaction .
[ lem : localexclusionlr ] for any @xmath9 , @xmath16 and @xmath616 a square , and with @xmath617 , we have that @xmath673 and @xmath674 \ge \frac{{e_{\mathrm{lr}}}(\alpha , \gamma(q))}{|q| } \biggl ( \int_q\varrho_\psi \ -1 \biggr)_{\!\!+},\ ] ] with @xmath675 where @xmath676 . for @xmath89 ,
the above bounds are valid with @xmath677 for all @xmath9 , where @xmath678 \to \r$ ] is a function defined below satisfying @xmath679 ( see figure [ fig : idealplots ] for both lower and upper bounds for @xmath397 ) . at ( 0,0 ) on the unit square as a function of @xmath680 $ ] , obtained by means of the projection method ( blue ) and temple ( yellow ) , as well as the upper and lower bounds given in ( [ eq : elrideal ] ) ( green ) .
right : a numerical lower bound to the energy @xmath677 on the unit square as a function of @xmath26.,title="fig : " ] ; at ( 7.2,.2 ) ; at ( 0,0 ) on the unit square as a function of @xmath680 $ ] , obtained by means of the projection method ( blue ) and temple ( yellow ) , as well as the upper and lower bounds given in ( [ eq : elrideal ] ) ( green ) .
right : a numerical lower bound to the energy @xmath677 on the unit square as a function of @xmath26.,title="fig : " ] ; at ( 7.15,.2 ) ; the tiny constant @xmath681 stems from theorem [ thm : long - range ] and again we expect that it could be replaced with @xmath140 or just slightly smaller ( recall remark [ rem : mainradialbound ] )
. note that the factor in front of the energy for @xmath89 in the limit @xmath31 is exactly the expected one from average - field theory , @xmath682 for suitable @xmath26 , however the bound is only linear ( and not quadratic ) in @xmath211 and hence only good for small enough boxes @xmath606 , resulting in a worse constant ( by a factor @xmath141 ) when applied below in the thermodynamic limit .
also note that the bounds involve @xmath683 and not @xmath684 or @xmath685 because there is a probability that more particles ( in fact all the way up to @xmath0 ) can be found on @xmath606 .
* ideal case . *
we begin with the more transparent case of @xmath89 , and note that we may set @xmath686 by scaling .
our starting point is the long - range magnetic interaction bound provided by theorem [ thm : long - range ] .
for the ideal case the theorem states that @xmath687 in order to convert this non - uniform potential to a uniform bound for the energy we take part of the kinetic energy and then apply either temple s inequality as in lemma [ lem : templesmearbound ] or a projection argument as in ( * ? ? ?
* lemma 7 ) or lemma [ lem:1dprojection ] . to this end
we take a fraction @xmath688 $ ] of the original kinetic energy for which we use the diamagnetic inequality and the identity @xmath689 and on the remaining fraction @xmath642 we use theorem [ thm : long - range ] .
we then obtain that - dependence in the first term if we are aiming for a bound which is convex in @xmath211 . ]
@xmath690 & \ge ( n-1)_+ \ , { e_{\mathrm{lr}}}(\alpha , 0 ) , \end{aligned}\ ] ] where @xmath691 and @xmath692 we then use the convexity in @xmath211 to obtain the corresponding bound for @xmath649 $ ] in terms of @xmath693 .
the upper bound @xmath694 is found simply by taking the trial state @xmath695 and then @xmath112 , carrying out the integration as below ( with @xmath696 ) .
we now wish to find a lower bound for the integral in @xmath697 , which then is to be maximized in @xmath636 .
this is equivalent to finding a lower bound for the ground - state energy of the schrdinger operator on @xmath698 @xmath699 however , to apply a projection bound or use temple s inequality requires that @xmath700 and @xmath701 , respectively .
as neither of these conditions are satisfied for our @xmath666 we use the fact that @xmath702 and thus truncating our potential will only lower the energy .
therefore we instead study the eigenvalue problem with @xmath666 replaced by the truncated potential @xmath703 defined in relative coordinates by @xmath704 \displaystyle{\frac{{{\ensuremath { \mathds 1 } } } _ { b_{\delta({\mathbf{x}})}}({\mathbf{r}})}{\hat\delta^2 } } , & \delta({\mathbf{x } } ) < \hat\delta \end{matrix}\right.\ ] ] ( in slightly more compact notation , @xmath705 ) . as @xmath706 , @xmath707
, it follows that also @xmath708 .
we proceed by calculating the expectation of @xmath703 and @xmath709 in the ground state @xmath710 of @xmath711 , as needed for the bounds . through a straightforward calculation
one finds that @xmath712 ^ 2}\int_{q_{\mathbf{x}}}\frac{1}{\delta({\mathbf{x}})^2}\ , d{\mathbf{r}}d{\mathbf{x}}+\int_{q\setminus[\hat\delta , 1-\hat\delta]^2}\int_{q_{\mathbf{x}}}\frac{1}{\hat\delta^2}\ , d{\mathbf{r}}d{\mathbf{x}}\biggr)\\ & = 2\pi \bigl(1 + 2\hat\delta^2- \frac{8\hat\delta}{3}\bigr ) , \end{aligned}\ ] ] and correspondingly for @xmath709 we obtain @xmath713 ^ 2}\int_{q_{\mathbf{x}}}\frac{1}{\delta({\mathbf{x}})^4}\ , d{\mathbf{r}}d{\mathbf{x}}+\int_{q\setminus[\hat\delta , 1-\hat\delta]^2}\int_{q_{\mathbf{x}}}\frac{1}{\hat\delta^4}\ , d{\mathbf{r}}d{\mathbf{x}}\biggr)\\ & = 4\pi \bigl(\frac{8}{3\hat\delta}+4\ln(2\hat\delta)-5\bigr).\end{aligned}\ ] ] choosing @xmath714 for some @xmath715 $ ] ( this normalization is convenient ) results in @xmath716 our considerations here have been for @xmath717 the unit square but also other domains @xmath102 could be of interest .
similar calculations when @xmath102 is the unit disk and @xmath718 give instead @xmath719 let @xmath290 denote the orthogonal projection onto the ground state @xmath720 , and let @xmath721
. then @xmath722 , and with @xmath723 the first non - zero neumann eigenvalue , @xmath724 arguing as in lemma [ lem:1dprojection ] , we for any @xmath725 obtain that @xmath726 the first of these operators is equal to @xmath727 , and we can control the second term by using that @xmath728 .
thus , for any @xmath729 we find that @xmath730 the last expression , seen as a function in @xmath731 , is piecewise linear and concave .
thus to obtain the largest linear minorant of this function it suffices to find the largest value attained at the right endpoint of our range of values @xmath731 , that is at @xmath732 . by the @xmath733 dependence of each of the two terms in the minimum this quantity
is seen to be maximal when the two terms are equal . solving this quadratic equation in @xmath733 and choosing @xmath734
we find that @xmath735 to obtain that @xmath736 we apply temple s inequality ( as in lemma [ lem : templesmearbound ] ) . in our current setting it yields that @xmath737 provided that @xmath738 .
we decrease the above quantity by throwing away positive terms and increasing the denominator of the last term yielding @xmath739 the positivity of denominator is then ensured if @xmath740 .
we can thus , for @xmath731 sufficiently small , choose @xmath741 for some @xmath742 to be fixed later . inserting this into our expression
we find that @xmath743 setting @xmath744 , @xmath745 , we obtain that @xmath746 and choosing @xmath747 yields @xmath748 inserting this into we have @xmath749 which completes the proof . * extended case .
* let , in the case that @xmath16 , @xmath750 denote the relative length scale of the interaction , @xmath751 , and note that we may again rescale everything so that @xmath686 .
we then proceed as above using projection , where the bound from theorem [ thm : long - range ] is replaced by @xmath752 where @xmath753 and @xmath754 is an additional parameter that we may optimize over , however we will in order to simplify the analysis take @xmath755 .
since @xmath756 maximally takes the value @xmath757 , the above expression is zero for @xmath758 .
for @xmath759 we can proceed by truncating to the , in @xmath750 , uniformly bounded potential @xmath760 with the support ( consisting of truncated relative annuli ) @xmath761 and therefore , since @xmath103 is monotonically decreasing in @xmath750 , @xmath762 also , using the coarea formula and that @xmath763 almost everywhere , we obtain that @xmath764 where in the last step we again used the monotonicity of @xmath115 , and @xmath765 where the lower bound is found by taylor expansion around @xmath766 .
thus , the corresponding projection bound for the operator @xmath767 reads @xmath768 we take @xmath769 and @xmath770 , and use that @xmath771 , to obtain the claimed bound @xmath772 with @xmath773 and @xmath774 .
again we may use the convexity in @xmath211 to obtain the corresponding bound for @xmath649 $ ] .
let us finally consider the homogeneous gas in the thermodynamic limit , i.e. @xmath0 particles confined to a large box ( square ) @xmath78 , where we shall take simultaneously @xmath775 and @xmath776 while keeping the density @xmath77 fixed .
the only dimensionless parameters are then the magnetic interaction strength @xmath9 and the relative interaction length scale ( magnetic filling ratio ) @xmath777 , also held fixed , so that in the limit the ground - state energy , @xmath778 per particle must for dimensional reasons be given by @xmath779 where @xmath780 is dimensionless .
we have that @xmath781 for all @xmath782 , corresponding to non - interacting bosons , and @xmath783 for ideal fermions in two dimensions due to the weyl asymptotics for the laplacian eigenvalues .
we also have a reflection - conjugation symmetry @xmath784 for all @xmath785 .
furthermore , in the dilute limit we should see a periodicity in the entire spectrum with respect to any shift in @xmath26 by an even integer , and in particular @xmath786 due to the gauge equivalence . on the other hand
, average - field theory suggests a linear dependence @xmath787 for arbitrary @xmath26 and large enough @xmath20 .
hence there must be some non - trivial interpolation between these regimes .
although the existence of the thermodynamic limit might be expected on physical grounds , as is indeed the case for bosons and fermions with reasonable scalar interactions ( see e.g. @xcite ) , we are not aware of any proof of it for anyons , whose interaction is long - range and magnetic instead of scalar .
furthermore , there is for anyons also a subtlety in the choice of boundary conditions , partly since topology plays an important role in the whole problem and therefore periodic b.c .
may seem problematic , and even in the case of a constant magnetic field we know that neumann and dirichlet b.c .
differ substantially ( cf . section [ sec : short - range ] and proposition [ prop : constant - field ] ) .
we shall therefore replace the limit with the @xmath788 and also stick to dirichlet b.c .
( ` hard - wall ' confined anyons ) in all that follows .
[ thm : gas ] let @xmath129 , where @xmath130 , denote the ground - state energy per particle and unit density of the extended anyon gas in the thermodynamic limit at fixed @xmath9 , @xmath16 and density @xmath131 where dirichlet boundary conditions have been imposed , that is @xmath789 then @xmath790 for some universal constant @xmath791 , with @xmath792 given in lemma [ lem : localexclusionsr ] , and @xmath793 in lemma [ lem : localexclusionlr ] .
furthermore , for any @xmath9 and with @xmath397 given in lemma [ lem : localexclusionlr ] , we have for the ideal gas that @xmath794 moreover , for any fixed @xmath795 we obtain in the dilute limit that @xmath796 while if @xmath797 is arbitrary but fixed , and @xmath798 then @xmath799 note that for the short - range part of the interaction , one can view the height of the potential compared to the average density as a dimensionless interaction strength , and that in the dilute limit with fixed @xmath800 we have that @xmath801 corresponding to a hard - core interaction . on the other hand , under the conditions in , @xmath802 and thus corresponding to a very weak soft - core interaction rather than a hard - core one in this regime .
we also note that average - field theory has indeed been proved to be correct for the trapped anyon gas in a certain almost - bosonic regime ; see @xcite . here
this corresponds to taking @xmath803 fixed , @xmath804 and @xmath805 with @xmath806 , in which case we have that @xmath807 and @xmath808 as @xmath775 .
however , the sense in which its validity is confirmed is that all the anyons become identically distributed subject to a self - consistent magnetic field , and it should be remarked that the constant @xmath809 that is predicted by the usual average - field approximation does not take such self - interactions fully into account and may ultimately be replaced by a larger one , at least in a particular limit @xcite .
let us begin with the universal bound for all @xmath785 .
we have a sequence of @xmath84 and squares @xmath78 with @xmath810 , and consider in each case an arbitrary function @xmath96 supported on @xmath811 .
let us again write @xmath812 & : = \langle \psi , \hat{t}_\alpha \psi \rangle = \kappa_1 \sum_{j=1}^n \int_{\r^{2n } } |d_j\psi|^2\ , d{\textup{x}}+ \kappa_2 \sum_{j=1}^n \int_{\r^{2n } } |d_j\psi|^2\ , d{\textup{x}}+ \kappa_3 \sum_{j=1}^n \int_{\r^{2n } } |d_j\psi|^2\ , d{\textup{x}}\\ & \ge \int_{\r^{2n } } \sum_{j=1}^n \biggl ( \kappa_1 \bigl|\nabla_j|\psi|\bigr|^2 + \kappa_2 \sum_{k \neq j } 2\pi|\alpha| \frac{{{\ensuremath { \mathds 1 } } } _ { b_r(0)}}{\pi r^2}({\mathbf{x}}_j - { \mathbf{x}}_k ) \ , |\psi|^2 + \kappa_3 |d_j\psi|^2 \biggr ) d{\textup{x}}.\end{aligned}\ ] ] take @xmath813 and @xmath814 , and a partition of @xmath803 into @xmath815 squares @xmath606 of equal size .
then , by the local exclusion principles of lemma [ lem : localexclusionsr ] and lemma [ lem : localexclusionlr ] , @xmath816 \ge n^{-1 } \sum_q t_q^{(\kappa/2 , \kappa/2 , 1-\kappa)}[\psi ] \\ & \ge n^{-1 } \sum_q |q|^{-1 } \bigl ( \frac{\kappa}{2 } { e_{\mathrm{sr}}}\bigl(\alpha , \gamma(q ) , { \textstyle \int_q \varrho_\psi}\bigr ) + ( 1-\kappa ) { e_{\mathrm{lr}}}\bigl(\alpha , \gamma(q)\bigr ) \bigr ) \biggl ( \int_q\varrho_\psi \ -1 \biggr)_{\!\!+ } \\ & \ge n^{-1}|q_0|^{-1 } m^2 \sum_q \biggl ( \int_q\varrho_\psi \ -1 \biggr)_{\!\!+ } \times \\ & \quad\times \left\{\begin{array}{ll } \frac{\kappa}{2 } |\alpha| \min\bigl\{(1-\gamma(q)^2/2)^{-1 } , k_\alpha/2\bigr\ } \bigl ( k_\alpha + 2|\alpha|\bigl ( -\ln ( \gamma(q)/ & \hspace{-10pt } { \sqrt{2 } } ) \bigr ) \bigr)^{-1 } \\[8pt ] \qquad + \ ( 1-\kappa ) \frac{\pi}{48 } g\bigl(c\alpha_n , 12\gamma(q)\bigr)^2 ( 1 - 12\gamma(q))_+^3 , & \hspace{10pt}\text{for}\ \gamma(q )
< \sqrt{2 } \\[9pt ] \kappa |\alpha|\gamma(q)^{-2 } \int_q \varrho_\psi , & \hspace{10pt}\text{for}\ \gamma(q ) \ge
\sqrt{2}. \end{array}\right.\end{aligned}\ ] ] note that @xmath817 and we are free to choose @xmath242 $ ] and the integer @xmath818 as we like .
we choose @xmath819 for suitable @xmath820 , so that @xmath821 . then for @xmath822 we have , using @xmath823 , that @xmath824 \ge \bar\varrho \mu^2 ( 1 - \mu^2)_+ \biggl ( & \frac{\kappa}{2 } |\alpha| \frac { \min\bigl\{(1-\mu^2\bar\gamma^2/2)_+^{-1 } , k_\alpha/2\bigr\ } } { k_\alpha + 2|\alpha|\bigl ( -\ln ( \mu\bar\gamma/{\sqrt{2 } } ) \bigr ) } \\ & + ( 1-\kappa ) \frac{\pi}{48 } g(c\alpha_n , 12\mu\bar\gamma)^2 ( 1 - 12\mu\bar\gamma)_+^3 \biggr).\end{aligned}\ ] ] on the other hand , if @xmath825 , we by using convexity , @xmath826 have that @xmath827 \ge \kappa|\alpha|\bar\varrho \bar\gamma^{-2 } ( \mu^{-2 } - 1)_+.\end{aligned}\ ] ] hence , in the case @xmath828 we can in the thermodynamic limit choose @xmath829 and @xmath830 in order to obtain that @xmath831 while for @xmath832 we choose , for simplicity , @xmath833 and @xmath834 obtaining that @xmath835 this proves the first part of the theorem with @xmath836 . in the ideal case @xmath89 , and hence @xmath837 , we take @xmath112 and @xmath838 in ( which means approximately 2 particles in each box ) to obtain from of lemma [ lem : localexclusionlr ] . the second bound in follows immediately from and the properties of @xmath115 , by setting @xmath112 and @xmath839 . for the first bound we set @xmath840 , @xmath841 and @xmath842 in and use the result @xcite of lieb and yngvason for the dilute repulsive bose gas in two dimensions .
we find for the ( bosonic , and therefore positive ) ground state @xmath843 of this expression , with fixed @xmath652 and @xmath320 , that @xmath844}{n\bar\varrho } & \ge \frac{1-\kappa}{n\bar\varrho } \int_{\r^{2n } } \biggl ( \sum_{j=1}^n \bigl|\nabla_j\psi_0\bigr|^2 + \sum_{j < k } w({\mathbf{x}}_j - { \mathbf{x}}_k ) \ , |\psi_0|^2 \biggr ) d{\textup{x}}\\ & = \frac{4\pi(1-\kappa)}{|{\ln a_r^2\bar\varrho}| } \bigl ( 1+o\bigl(|{\ln a_r^2\bar\varrho}|^{-1/5}\bigr ) \bigr ) = \frac{2\pi(1-\kappa)}{k'_{\alpha , \kappa } - \ln\bar\gamma } \bigl ( 1+o\bigl((k'_{\alpha , \kappa } - \ln\bar\gamma)^{-1/5}\bigr ) \bigr ) , \end{aligned}\ ] ] where we used that the pair potential @xmath845 has scattering length ( cf . ) @xmath846 with @xmath847 hence for any @xmath320 and @xmath848 we , by setting @xmath849 and then taking the limit @xmath850 , obtain that @xmath851 so for each fixed @xmath795 @xmath852 to obtain the bound for the soft - core regime we follow @xcite . again
we partition @xmath803 into @xmath815 squares @xmath606 of equal size , and let @xmath853 . with @xmath688
$ ] we then have that @xmath854 & \geq n^{-1}\sum_q t_q^{(\kappa , 1-\kappa , 0)}[\psi ] \geq n^{-1}\sum_q\sum_{n\geq0 } e_n^{(\kappa , 1-\kappa , 0)}(|q|)p_n(\psi;q).\end{aligned}\ ] ] set @xmath855 , i.e. @xmath856 is the fraction of cells @xmath606 containing precisely @xmath211 particles , then @xmath857 rearranging the sum and from now on suppressing the weight @xmath858 we find that @xmath859 & \geq \frac{1}{\bar\varrho\ell^2 } \sum_{n\geq 0}e_n(|q| ) c_n , \end{aligned}\ ] ] which is precisely the starting point of the argument in @xcite .
fix @xmath860 .
since the energy is superadditive , @xmath861 , we for all @xmath862 have that @xmath863 applying lemma [ lem : templesmearbound ] yields @xmath864 where @xmath865 if the expression in the last denominator is positive and @xmath866 otherwise .
if instead @xmath867 we use that @xmath868 is decreasing in @xmath211 to find @xmath869 splitting the sum into two we thus find that @xmath870 we wish to minimize @xmath871 set @xmath872 by convexity is then larger than @xmath873 if @xmath874 and @xmath875 this is minimized at @xmath876 , where it is equal to @xmath877 .
thus by choosing @xmath878 we have shown that @xmath879 \geq 2\pi |\alpha|\bar\varrho\bigl(1-\frac{1}{\bar\varrho\ell^2}\bigr)_{\!+ } k(4\bar\varrho\ell^2 , \ell),\ ] ] and hence , upon taking the thermodynamic limit @xmath880 with all the other parameters kept fixed , @xmath881 as long as @xmath882 .
given @xmath883 , let us choose @xmath849 and also demand that @xmath884 , @xmath885 , @xmath886 , and @xmath887 .
we therefore choose @xmath888 and then find that , together with the requirement on @xmath26 and @xmath889 which implies @xmath890 , all conditions above are satisfied , and the error terms in are of order @xmath889 or higher .
[ prop : besselprimezerobounds ] for @xmath891 we let @xmath37 denote the first positive zero of the derivative of the bessel function @xmath38 .
then we have that @xmath892 a proof of the above proposition and much more refined bounds for the zeros of bessel functions and their derivatives can be found in @xcite . for completeness
we provide an elementary proof which covers our needs . by a standard variational argument
it can be shown that @xmath893 where the infimum is taken over all @xmath894 , rdr)$ ] and is attained by @xmath895 . for @xmath891 and @xmath894 , rdr)$ ] with @xmath896
we obtain using hlder s inequality that @xmath897\\ & \leq 2 \biggl(\int_0^t|u'(r)|^2r\ , dr \biggr)^{1/2}\biggl(\int_0^t|u(r)|^2r^{-1}\ , dr\biggr)^{1/2}\\ & = \frac{2}{\nu } \biggl(\int_0^t|u'(r)|^2r\ , dr \biggr)^{1/2}\biggl(\nu^2 \int_0^t|u(r)|^2r^{-1}\ , dr\biggr)^{1/2}.\end{aligned}\ ] ] through an application of young s inequality we then find @xmath898 and integrating both sides in @xmath731 over @xmath899 against @xmath900 yields @xmath901 which implies that @xmath902 taking the infimum over all functions @xmath903 , rdr)$ ] such that @xmath896 , in particular this includes @xmath38 , we see that @xmath904 which completes the proof of the lower bound . to obtain the upper bound , simply take @xmath905 in the variational quotient above . in the case of @xmath17-extended anyons
our bounds result in studying the behavior of solutions to a bessel - type eigenvalue equation of order @xmath623 with neumann boundary conditions on the interval @xmath906 , for some @xmath907 .
thus it is of interest for us to understand the behavior of the lowest eigenvalue of such an equation in both parameters @xmath750 and @xmath623 .
[ prop : besseleigenvaluebounds ] given @xmath908 and @xmath909 , let @xmath910 , where @xmath106 denotes the first positive solution to the eigenvalue equation @xmath911 with the neumann boundary conditions @xmath912 .
then , for fixed @xmath750 , @xmath913 is a monotonically increasing function in @xmath623 .
also , for fixed @xmath623 , @xmath913 is a monotonically decreasing function of @xmath750 , and satisfies @xmath914 moreover , we have that @xmath915 and @xmath916 .
that @xmath103 is monotonically increasing in @xmath623 is clear from the variational characterization of @xmath106 , @xmath917 it is well known that the solution of the above differential equation is given by a linear combination of the bessel functions @xmath918 and @xmath919 . only if @xmath750 were zero could we exclude the bessel function of the second kind since it fails to be in @xmath920,r dr)$ ] and thus can not be a solution
. thus the problem reduces to finding the smallest @xmath921 such that the system @xmath922 admits a non - trivial solution , which is equivalent to the determinant equation @xmath923 assuming that @xmath924 is smaller than the first zero of @xmath925 ( this will be seen to be true once we find our solution ) we can equivalently solve the equation @xmath926 letting @xmath927 we find that @xmath928 where we used that @xmath38 and @xmath929 satisfy the bessel equation and the well - known identity @xmath930 ; see e.g. ( * ? ? ?
10.5.2 ) .
thus @xmath931 is strictly increasing on @xmath932 and decreasing after that .
we also know that @xmath933 .
but then it is clear that the graph of @xmath931 and that of its dilation @xmath934 must intersect between @xmath935 and the minimum of @xmath936 and @xmath937 ( compare figure [ fig : besseldilationequationplot ] ) , and as this solution is less than the first zero of @xmath925 the assumption above is seen to be true .
moreover , as @xmath938 we see that the solution @xmath939 tends to the zero @xmath37 and if instead @xmath940 it tends to the max point @xmath623 . at ( 0,0 ) ; at ( -0.035,0.4 ) ; at ( 3.68,0.4 ) ; at ( 4.9,0.35 ) ; at ( 6.73,0.35 ) ; at ( 7.3,0.32 ) ; by the above geometric considerations we can conclude that for @xmath745 and @xmath891 we have that @xmath106 , the smallest positive eigenvalue of , satisfies @xmath941,\ ] ] and is monotonically decreasing in @xmath750 .
[ prop : concavity ] for any @xmath394 the function @xmath441 given by is concave on its support intersected with @xmath425 . if in addition @xmath403 the function is concave on its full support @xmath404 $ ] . without
loss of generality we may , and do , assume that @xmath450 .
the proof is then a straightforward computation .
we begin with assuming that @xmath942 . for such @xmath417
the function @xmath441 is @xmath943 on @xmath944 $ ] which reduces the statement to proving that @xmath945 in this region .
calculating this derivative one finds @xmath946 and clearly the overall sign is determined by that of the polynomial in the denominator @xmath947 we need to prove that @xmath948 for @xmath949 in the triangular region given by @xmath456 where @xmath950 .
we first check the statement on the boundary of the region : @xmath951 thus all that remains is to check that we have no stationary points for @xmath126 in the interior of the region . calculating the derivative in @xmath240
one finds that @xmath952 as this is a quadratic polynomial in @xmath953 we can solve the equation @xmath954 and find that there are no solutions in our region .
this completes the proof of the claim in the case @xmath942 . in the case
@xmath955 we wish to prove that @xmath441 is concave on @xmath956 $ ] .
it is here convenient to study the problem in the variables @xmath417 and @xmath957 , and letting @xmath958.\ ] ] differentiating twice in @xmath959 we find that @xmath960 where @xmath961 as before the sign of is determined by that of the polynomial @xmath290 . if we can prove that @xmath962 for all @xmath963 and @xmath964 the claim follows . to this end
we proceed as above .
the values of @xmath290 on the boundaries of this region are ( in the same manner as before ) readily checked to be negative : @xmath965 what remains is to check stationary points in the interior .
for this polynomial solving either of the equations @xmath966 or @xmath967 is slightly harder .
however , since certain terms cancel one can instead solve the equation @xmath968 and the solutions are @xmath969 and @xmath970 .
the third solution is the only one contained within our region .
evaluating the derivative at this solution we obtain @xmath971 and since this is non - zero in the interior of our domain the proof is complete .
height 2pt depth -1.6pt width 23pt , _ stability of matter _ , in statistical physics , phase transitions and superfluidity , brandeis university summer institute in theoretical physics 1966 , gordon and breach , new york , 1968 , pp .
179239 .
, _ lectures on the coulomb stability problem _ , in statistical mechanics and mathematical problems , battelle rencontres , seattle , wa , 1971 , vol .
20 of lecture notes in phys . , springer - verlag , berlin , heidelberg , 1973 , pp .
114135 .
height 2pt depth -1.6pt width 23pt , _ inequalities for the moments of the eigenvalues of the schrdinger hamiltonian and their relation to sobolev inequalities _ , studies in mathematical physics , princeton univ . press , 1976 , pp .
269303 .
height 2pt depth -1.6pt width 23pt , _ the ground state energy of a dilute bose gas _ , in differential equations and mathematical physics ( birmingham , al , 1999 ) , vol .
16 of ams / ip stud . adv .
soc . , providence , ri , 2000 , pp .
295306 .
height 2pt depth -1.6pt width 23pt , _ the ground state energy of a dilute two - dimensional bose gas _
, j. statist . phys .
, 103 ( 2001 ) , pp .
509526 . special issue dedicated to the memory of joaquin m. luttinger . ,
_ anyons _ , in topological aspects of low dimensional systems , a. comtet , t. jolicur , s. ouvry , and f. david , eds .
69 of les houches - ecole dete de physique theorique , ( springer - verlag , berlin , germany ) , 1999 , pp .
265413 . , _ structure des systmes dynamiques _ , matrises de mathmatiques , dunod , paris , 1970 .
english translation by r. h. cushman and g. m. tuynman , progress in mathematics , 149 , birkhuser boston inc . ,
boston , ma , 1997 . | we introduce a rigorous approach to the many - body spectral theory of extended anyons , i.e. quantum particles confined to two dimensions that interact via attached magnetic fluxes of finite extent .
our main results are many - body magnetic hardy inequalities and local exclusion principles for these particles , leading to estimates for the ground - state energy of the anyon gas over the full range of the parameters .
this brings out further non - trivial aspects in the dependence on the anyonic statistics parameter , and also gives improvements in the ideal ( non - extended ) case . |
to extract an equation of state ( eos ) for nuclear matter from finite nuclear system , is one of most important task given to nuclear physics .
the eos is a very important topic to connect the nuclear physics to the astrophysical objects such as nucleosynthesis , neutron star , and so on .
the eos is often expressed in the expansion around symmetric matter , as follows .
@xmath12&=&{\cal e}[\rho,\delta=0 ] + { \cal s}[\rho,\delta ] + { \cal o}[\delta^4 ] , \nonumber \\ { \cal e}[\rho]&\equiv & { \cal e}_0 + \frac{k_\infty}{2}\varrho^2 + \frac{q_0}{6}\varrho^3 + \cdots , \nonumber \\ { \cal s}[\rho,\delta]&\equiv & \left ( j + l\varrho + \frac{k_{\rm sym}}{2}\varrho^2 + \cdots \right)\delta^2 , \label{eos}\end{aligned}\ ] ] where @xmath13 is an asymmetric parameter which separates a symmetric matter ( sm ) eos @xmath14 and symmetry energy @xmath15 .
they are expressed in @xmath16 , which is an expansion around the nuclear saturation density @xmath17 at which the pressure of nuclear matter is zero . when the eos is expanded as eq.([eos ] ) , there appear characteristic parameters : the binding energy per nucleon @xmath18 , the incompressibility of sm @xmath4 , the skewness parameter of sm @xmath19 , the symmetry energy @xmath20 , the slope parameter @xmath21 and the symmetry incompressibility @xmath11 , and so on .
the @xmath18 and @xmath20 are respectively equal to the @xmath14 and @xmath15 at @xmath17 .
the other parameters can be obtained from the density derivation of @xmath14 and @xmath15 as follows .
@xmath22 & & q_0 = 27 \rho_0 ^ 3 \left . \frac{\partial^3 { \cal e}}{\partial \rho^3 } \right|_{\rho=\rho_0},\end{aligned}\ ] ] @xmath23 & & k_{\rm sym } = 9 \rho_0 ^ 2 \left . \frac{\partial^2 { \cal s}}{\partial \rho^2 } \right|_{\rho=\rho_0}. \end{aligned}\ ] ] thus the eos parameters can be easily obtained at when the energy density functional ( edf ) @xmath24 $ ] is chosen .
nucleus is a too much small system to extrapolate directly to the infinite nuclear matter @xcite .
therefore , we evaluate the eos with the help of effective interaction ( skyrme , gogny , relativistic mean field ) @xcite . however , it is not easy to determine acceptable values in nuclear structure and in astrophysics , at same time .
currently also still , a study of the eos using effective interactions has been progressed theoretically and also experimentally @xcite .
the uncertainties of @xmath18 and @xmath20 are expected to be small : @xmath25 16 mev which has appeared in the bethe - weiz@xmath26cker mass formula , and @xmath20=32@xmath93 mev @xcite .
the @xmath4 does not have so large uncertainty due to the consistency among experiments and theoretical prediction : @xmath4=230@xmath930 mev @xcite .
the slope parameter @xmath21 which will be strongly related to nuclear dipole mode , has been well studied from the many points of view , for instance the relation among the neutron - skin thickness @xcite , pygmy dipole resonance @xcite , giant dipole resonance ( gdr ) @xcite and polarizability @xcite , although it has been yet floated : @xmath21=58@xmath918 mev @xcite .
these untiring studies narrow downs the range of the eos parameters , however the @xmath11 especially has a large range of values . basically , it is difficult to connect between eos parameters and experimental values directly
, therefore we usually take the procedure : to search an edf to reproduce experiments and then to extract the eos parameters from the edf . in a present work ,
we evaluate incompressibility parameters from the isoscalar giant monopole resonance ( isgmr ) of finite nuclear system . in order to estimate the parameters independently of the speciality of each nucleus
, we use the finite nuclear incompressibility @xmath7-expansion to analyse them , although we can compare the isgmrs in theory and in experiments .
if the energy of isgmr @xmath27 is represented in the root mean square radius of the nucleus and the gmr can be regarded as a single phonon mode , @xmath27 is written in @xmath28 where @xmath7 is defined as an incompressibility of finite nuclear system @xcite .
although the @xmath7 can not directly equal to @xmath4 at a limit of @xmath29 , it will bring the relation between the parameters of finite and infinite system .
the @xmath7 is expanded around @xmath4 as @xmath30 when incompressibility parameters are extracted from the experiments , the expansion ( [ kaex ] ) is often used @xcite .
the surface and coulomb term ( @xmath31 , @xmath6 ) are estimated by infinite eos parameters in previous studies , which will be confirmed how available in this work .
the isospin term @xmath8 has not been clear , which is a most important key to know the information of nuclear matter incompressibility using eq.([kaex ] ) .
this paper is organized as follows . in sec.[sec : method ] , we present the methods to calculate ism mode and to evaluate the centroid energy of isgmr @xmath32 . in sec.[sec : result ] , at first , we show partially the strength functions of @xmath33 nuclide to check the behaviour of them including the results of deformed nuclei . after that
, we extract the eos parameters of finite nuclear system by the chi - square fitting with eq.([kaex ] ) , from the @xmath7 evaluated with eq.([egmr ] ) and @xmath32 .
in particular , we investigate @xmath33 nuclide for @xmath31 and @xmath6 , and isobar nuclide for @xmath8 .
furthermore , to confirm the expansion itself and with our coefficients , we compare them with the experimental @xmath27 and calculated @xmath32 for sn isotopes . in sec.[sec : discuss ] , we also compare the experimental measurements for a whole mass region with @xmath3=24@xmath34238 , in order to narrow down the candidates of effective interaction .
and we mention the mass dependence of @xmath27 in experiments .
the present work is summarized in sec.[sec : con ] .
[ sec : introduction ]
to access the gmr , we apply the canonical - basis time - dependent hartree - fock - bogoliubov ( cb - tdhfb ) theory @xcite in three - dimensional ( 3d ) coordinate space which can be successfully applied to the study of the dipole @xcite and quadrupole @xcite modes of many isotopes , systematically .
the cb - tdhfb can describe the dynamical effects of pairing correlation in fully self - consistently .
the cb - tdhfb equations are derived from the full tdhfb equation represented in the canonical basis @xmath35 which diagonalize a density matrix , and by assuming the diagonal form of pairing functional .
the cb - tdhfb equations compose the time - evolution equations for the canonical pair @xmath36 , its occupation probability @xmath37 and pair probability @xmath38 , @xmath39 where the phase of canonical basis is chosen as @xmath40 , and the @xmath41 and @xmath42 are the single - particle hamiltonian and the gap energy , respectively . due to the appearance of deformed ground state in our subjective nuclide
, we should choose a flexible calculation space .
we use the 3d cartesian coordinate - space representation for canonical basis , @xmath43 with spin @xmath44 .
the condition of calculation space is discretized in a square mesh of 1.0 fm inside of a sphere of radius 15 fm for all nuclide in present work . to apply our method to systematic investigation
, we choose the skyrme edf to @xmath45-channel and the simple pairing functional form to @xmath46(@xmath47)-channel : @xmath48 where @xmath49 is constant in real - time evolution , as same as ref .
our choice of five skyrme parameter sets is sgii@xcite , skm@xmath0@xcite , sly4@xcite , skt3@xcite and ski3@xcite .
the reason to choose them is not only their usefulness , also corresponds to the limitation of @xmath4=230@xmath930 mev indicated in ref.@xcite [ sec : method ] in order to induce monopole responses , we add a weak instantaneous external field @xmath50 to initial states of the time evolution . here
the isoscalar monopole operator acting on nucleus is given as @xmath51 .
the amplitude of the external field is so chosen to be a small number @xmath52 @xmath53 to guarantee the linearity .
the strength function @xmath54 can be obtained through the fourier transformation of @xmath55 : @xmath56 e^{i(e+i\gamma/2)t } dt , \nonumber \\\end{aligned}\ ] ] where @xmath57 and @xmath58 are the ground and excited states , respectively .
@xmath59 is a smoothing parameter set to 1 mev for whole nuclide in present .
[ sec : cb - tdhfb ] we need a procedure to compute the mean energy of gr , in common among subjective nuclide . to compute the centroid energy @xmath60 of gr , we use @xmath61 , although there are some evaluations ( @xmath62 or @xmath63 ) .
the @xmath61 is computed as follows in present work .
@xmath64 where @xmath65 and @xmath66 are an upper and under cut - off energy respectively .
they should be decided with more carefully , because the @xmath32 is sensitive to them . in this work ,
we decide them as : @xmath67 and @xmath68 mev , where @xmath69 . the empirical formula @xmath70 is found in a droplet model @xcite .
purposely we chose this way , because eq.([egmr ] ) to relate the @xmath7 with nuclear response , which is based on the one phonon picture in other words the gmr is assumed as one mode .
[ sec : egmr ]
we evaluate the mean energy of gmr using eq.([bgmr ] ) , and from them the finite incompressibility @xmath7 is also evaluated according to eq.([egmr ] ) with using calculated @xmath71 in table [ tab : gs1 ] . to determine the expression parameters of @xmath7 according to eq.([kaex ] ) , we proceed a following way step by step .
first , we determine the @xmath72 using the results of @xmath33 nuclide in which the isospin term @xmath8 does not contribute to @xmath7 .
second , we determine the @xmath8 while using the @xmath31 and @xmath6 fixed in the first step . to obtain the @xmath8 for each skyrme interaction ,
we analyse isobar nuclide for both @xmath73100 and 132 at same time .
lastly , to confirm that the expansion of @xmath7 with determined parameters reproduces the results of sn isotopes , we calculate the centroid energies @xmath27 according to eq.([egmr ] ) with @xmath7 and @xmath71 in table [ tab : gs2 ] , and compare them with actually calculated @xmath32s . [ sec : result ] ism strength functions of even - even @xmath33 nuclide from @xmath74ca to @xmath75sn with skm@xmath0 are shown in fig.[fig:1 ]
. chain , dashed , doted and thick lines show the results of @xmath7620 , 30 , 40 and 50 , respectively .
we can see the broad strength distribution of @xmath74ca in heigh energy around 21 mev .
the distribution becomes localized and its centre shifts to low energy , as mass number increase . in these strength ,
split distributions can be seen in thin lines which are corresponding to @xmath77cr and @xmath78kr .
the split is caused by the coupling monopole with quadrupole excitations due to the deformation .
typically the quadrupole gr appear in lower energy than gmr , thus the @xmath32 of a well deformed nucleus shifts to low energy .
there are some strengths in vicinity of zero energy , which corresponds to numerical spurious mode due to the detail of mesh size and of time step .
they are excluded from the estimation of @xmath32 in eq.([bgmr ] ) .
figure [ fig:2 ] shows the @xmath32 for @xmath33 nuclide with five skyrme parameters .
filled symbols means results of spherical nuclei or the nuclei which have small deformation ( @xmath79 ) , open ones means those in deformed nuclei . over @xmath80 ,
the trend and relation among the results of each skyrme parameter become clarified .
the behaviour of @xmath32 in deformed nuclei diverges from the trend of spherical .
the order of @xmath32 can be almost corresponded to the order of @xmath81 magnitude ( refer to tab.[tab : gs3 ] ) .
figure [ fig:3 ] shows @xmath7 obtained by eq.([egmr ] ) .
same as fig.[fig:2 ] , the filed and open symbols correspond to the results of spherical and deformed nuclei , but we use the star symbol for double magic ( dm ) nuclei ( @xmath74ca , @xmath82ni , @xmath75sn ) .
the behaviour of @xmath7 well corresponds to that of @xmath32 , thus the trend and the relation among interactions are stable over @xmath80 , and the results of deformed nuclei clearly have difference from spherical nuclei . to obtain the expansion coefficients ( @xmath31 , @xmath8 , @xmath6 ) , we analyse our results according to the @xmath7 expansion in eq.([kaex ] ) .
the results of deformed nucleus are excluded from our analysis , because they have clearly a different trend from those of spherical nuclide .
if they can be included into the analysis , we will need the way to separate quadrupole and monopole modes . here
, two cases are considered : ( i ) excluding dm nuclei and ( ii ) including them .
our purpose is to extract nuclear matter properties from nucleus .
the effects due to the special nuclear structure such as a modes coupling in deformed nuclei , should be excluded from the matter property analysis . as mentioned in ref.@xcite
, the magicity effects in the incompressibility will appear , which should be confirmed in the comparison the ( i ) and ( ii ) . in this section ,
we fix the @xmath31 and @xmath6 expansion coefficients which are listed in the table [ tab : gs3 ] .
the surface term @xmath31 is often estimated as an opposite sign of @xmath4 @xcite , and the coulomb term @xmath6 is estimated in ref.@xcite as @xmath83 our @xmath84 are larger than @xmath85 , which have been mentioned already in ref.@xcite . although an assumption @xmath86 is sometimes used for non - relativistic models in previous analyses @xcite , it is not suitable actually and the difference over 30% from it might cause a serious missing in nuclear property . in both ( i ) and ( ii ) , @xmath6 closes to the @xmath87 .
@xmath31 and @xmath6 in ( ii ) are a little weaker and stronger than those of ( i ) , respectively in most interaction - cases .
the effect of dm on @xmath7 is regarded as small as shown in fig.[fig:3 ] , although small kinks appear at @xmath82ni .
the nuclear magicity is not so sensitive to the @xmath7 excluding light dm nuclei .
.root mean square radius @xmath88 [ fm ] and absolute value of quadrupole deformation parameter @xmath89 in ground state for @xmath33 and @xmath90 nuclide . [
cols="^,^,^,^,^,^,^,^,^,^,^,^,^,^,^ " , ] [ sec : nez ] in this section , we determine the isospin term @xmath8 from isobar nuclide @xmath3=100 and 132 , while using @xmath31 and @xmath6 fixed in previous section . in same as sec .
[ sec : nez ] , we exclude the deformed nuclei from the analysis .
figure [ fig:4 ] shows the @xmath7 of the spherical isobars with @xmath3=100 and 132 with respect to isospin asymmetry @xmath91 , in which the vertical chain line separates @xmath3=100 and 132 .
the root mean square radii and quadrupole deformations of the isobars are listed in table [ tab : gs1 ] .
the @xmath7 in fig.[fig:4 ] has a parabolic shape in @xmath91 which corresponds to the expansion eq.([kaex ] ) , however the centre of the parabolic function is not always at @xmath1=@xmath2 . in this analysis
, we also consider the two cases : ( i ) with dm and ( ii ) without dm .
the isospin term @xmath8 obtained in the cases are listed in table [ tab : gs3 ] .
the effects of dm are not so large also in @xmath8 excluding the results of ski3 .
when we exclude the result of dm from ski3 results , the points which can be used in the analysis are only four , therefore the analysis ambiguity becomes large . to determine the @xmath8 , the number of isobar nuclide is essential . in several papers , an isospin dependence of incompressibility for nuclear matter
is estimated at saturation density @xmath17 with a small isospin asymmetry @xcite , which can be written in @xmath92 the @xmath10 can not be regarded as the finite incompressibility @xmath8 , which is mentioned also in ref.@xcite , although the strong correlation between @xmath8 and @xmath10 can be expected naively .
the @xmath10 is often used to expand the @xmath8 : @xmath93 .
the @xmath94 could not be estimated because the number of isobar - chain sample is only two .
the absolute value of @xmath8 is usual smaller than @xmath10 in present work and the correlation among them does not seem simple . [
sec : isobar ] we obtain the expansion coefficients of @xmath7 in previous sections . to confirm the coefficients and the expansion eq.([kaex ] ) itself , we compare the @xmath27 in eq.([egmr ] ) with the coefficients ( @xmath95 ) , in the experiment and @xmath96 by the linear response calculation with eq.([bgmr ] ) , for sn isotopes ( @xmath3=100 - 132 ) .
figure [ fig:5 ] shows the @xmath27 for sn isotopes .
solid lines and filled symbols mean the results directly calculated with eq.([bgmr ] ) , dashed lines correspond to the @xmath27 in eq.([egmr ] ) with the coefficients of case ( i ) for each interaction , and open square symbols are experimental data at rcnp@xcite .
furthermore , to show the pairing effects we add the @xmath96 with tdhf only for skm@xmath0 , which are symbolized by open circles and dotted line .
the dashed lines well reproduce whole @xmath32 of sn isotopes within a smaller than 0.3 mev .
it means the expansion eq.([kaex ] ) is an effective procedure and the coefficients are suitable .
the comparison our results and he experimental data may recommend skm@xmath0 and sgii parameters as a candidate of the `` _ answer _ '' .
the pairing effects to the trend on sn isotope isgmr can be discussed in the comparison between our results obtained by tdhf ( open ) and by cb - tdhfb ( filled ) with skm@xmath0 .
the small difference between them appears in whole isotopes , which can be expected due to the small deformation of hf ground states .
while the effect to soften eos slightly in a surface - type pairing functional was reported @xcite , our results indicate the opposite effects , however whose mechanism is different from the previous studies .
the hf ground states in sn isotopes have some deformed aspects as table [ tab : gs2 ] . as the explanation in sec .
[ sec : nez ] , the centroid energy of isgmr is estimated at lower than that of spherical nucleus due to the coupling with other modes , while using the summation analysis such as eq.([bgmr ] ) .
cccccccccc & sgii & & & & sly4 & skt3 & ski3 + + & @xmath97 & & @xmath97 & @xmath98 & @xmath99 & & @xmath97 & @xmath97 & @xmath97 + @xmath100sn & 4.39 & & 4.405 & 4.401 & 0.04 & & 4.42 & 4.38 & 4.40 + @xmath101sn & 4.43 & & 4.434 & 4.426 & 0.05 & & 4.45 & 4.40 & 4.43 + @xmath102sn & 4.46 & & 4.462 & 4.448 & 0.00 & & 4.48 & 4.43 & 4.47 + @xmath103sn & 4.48 & & 4.491 & 4.488 & 0.07 & & 4.51 & 4.46 & 4.50 + @xmath104sn & 4.51 & & 4.521 & 4.527 & 0.10 & & 4.54 & 4.49 & 4.53 + @xmath105sn & 4.54 & & 4.551 & 4.558 & 0.10 & & 4.57 & 4.52 & 4.56 + @xmath106sn & 4.57 & & 4.581 & 4.596 & 0.11 & & 4.60 & 4.54 & 4.59 + @xmath107sn & 4.59 & & 4.609 & 4.632 & 0.15 & & 4.62 & 4.57 & 4.61 + @xmath108sn & 4.62 & & 4.636 & 4.699 & 0.28 & & 4.65 & 4.60 & 4.64 + @xmath109sn & 4.64 & & 4.662 & 4.677 & 0.17 & & 4.67 & 4.63 & 4.67 + @xmath110sn & 4.67 & & 4.688 & 4.703 & 0.14 & & 4.70 & 4.66 & 4.70 + @xmath111sn & 4.69 & & 4.712 & 4.724 & 0.10 & & 4.72 & 4.68 & 4.72 + @xmath112sn & 4.71 & & 4.736 & 4.743 & 0.00 & & 4.74 & 4.71 & 4.75 + @xmath113sn & 4.74 & & 4.759 & 4.766 & 0.06 & & 4.76 & 4.74 & 4.77 + @xmath114sn & 4.76 & & 4.781 & 4.784 & 0.04 & & 4.79 & 4.76 & 4.80 + cccccrcccccccccc & & & & & + + int . &
@xmath17 & @xmath4 & @xmath19 & @xmath21 & @xmath11 & @xmath115 & @xmath87 & & @xmath31 & @xmath8 & @xmath6 & & @xmath31 & @xmath8 & @xmath6 + sgii & .1583 & 214.6 & -380.9 & 37.63 & -145.9 & -304.9 & -4.69 & & -273.7 & -295.9 & -4.52 & & -260.4 & -283.5 & -5.10 + skm@xmath0 & .1603 & 216.6 & -386.1 & 45.78 & -155.9 & -349.0 & -4.70 & & -285.3 & -282.0 & -4.76 & & -276.7 & -285.0 & -5.09 + sly4 & .1595 & 229.9 & -363.1 & 45.96 & -119.7 & -322.9 & -4.85 & & -318.4 & -314.3 & -4.73 & & -308.3 & -305.4 & -5.13 + skt3 & .1610 & 235.7 & -382.7 & 55.31 & -132.1 & -374.1 & -4.83 & & -326.1 & -305.1 & -4.54 & & -325.0 & -305.2 & -4.51 + ski3 & .1598 & 258.2 & -303.8 & 100.5 & 73.03 & -411.8 & -5.13 & & -372.6 & -352.0 & -4.75 & & -364.8 & -332.6 & -4.99 + [ sec : tin - iso ]
we obtain the expansion coefficients ( @xmath95 ) through the sec.[sec : nez ] and [ sec : isobar ] , and confirm them in sn isotopes .
skm@xmath0 and sgii parameters might be likely candidates in sec.[sec : tin - iso ] , however it is not clear that they can approach to more heavier system , at least in present results .
the many gmr distributions for whole mass region have been measured in the past , which are summarised in ref.@xcite .
we extend the mass region to apply eq.([egmr ] ) , and attempt to narrow down the parameters .
figure [ fig:6 ] shows experimental @xmath27s ( square ) which are measured at texas a&m , grennoble , groningen @xcite and rcnp@xcite , and the centroid energy ( solid line ) by eq.([egmr ] ) for each skyrme parameter but with @xmath71 expressed as @xmath116= 0.895@xmath117 0.321 fm which is obtained by fitting systematic mean radii calculated in skm@xmath0 .
the experimental data cover a wide mass region @xmath3=24@xmath34238 with @xmath2=12@xmath3492 . in @xmath11850 mass region ,
the mass dependence of experiment values is not clear apparently . over @xmath3=100
, they make a visible trend , though some values around @xmath3=90 deviate from the trend .
whole mass region , the solid lines keep their magnitude order which is same one in fig.[fig:5 ] , thus namely ski3 , skt3 , sly4 , sgii and skm@xmath0 in decreasing order . for ski3 , the results overestimate experiments . for sgii and skm@xmath0
they underestimate over @xmath3=160 , while they can well reproduce those around @xmath3=110 .
if we emphasize the agreement with a heavy system , sgii and skm@xmath0 can not reach to the conclusion as the best choice .
ski3 should be excluded from the candidate , due to the missing over @xmath3=100 . from the agreements with @xmath27 over @xmath3=140 ,
sly4 and skt3 are good candidates .
therefore our results indicate that the range of @xmath8 is -305@xmath910 mev , and the incompressibility parameters for infinite system corresponding to the @xmath8 are @xmath10=-340@xmath935 , @xmath4=225@xmath911 and @xmath11=-138@xmath918 mev , which are expected from sgii , skm@xmath0 , sly4 and skt3 parameters . in this paper ,
our analysis has constructed within the assumption eq.([egmr ] ) and ( [ kaex ] ) . as fig.[fig:6 ] shown that , the expansion eq.([kaex ] ) well work to connect between the infinite nuclear properties and finite nuclear excitation mode in eq.([egmr ] ) . however , the mass dependence of @xmath27 in experiment is apparently different from that scaled by @xmath119 , which has been mentioned also in ref.@xcite .
the small panel in fig.[fig:6 ] shows the fitting experimental data for @xmath3=90@xmath34238 , by @xmath119 ( dashed ) and by @xmath120 ( solid ) .
the coefficient of @xmath119 is consistent with the ordinary value @xcite , but it underestimates experiment in heavy mass region .
the solid line reproduces experimental data in a whole mass region .
the @xmath119 and @xmath121 scaling are well known on the gdr as steinwedel - jensen ( sj ) and goldhaber - teller ( gt ) model @xcite , respectively in which the dipole moments are caused by polarization and by proton - neutron displacement .
if we adapt the models to gmr , the sj and gt - model will respectively correspond to the density wave oscillation keeping nuclear size , and to the vibration between expansion and contraction of the size .
[ sec : discuss ]
we have performed a systematic investigation of the isgmr to extract the isospin dependent compression eos parameter from finite nuclear system , using the linear response calculation with cb - tdhfb represented in 3d coordinate space .
the expansion coefficients ( @xmath31 , @xmath8 , @xmath6 ) of finite incompressibility @xmath7 are determined from @xmath1=@xmath2 and isobar @xmath3=100 , 132 nuclide for each skyrme interaction ( sgii , skm@xmath0 , sly4 , skt3 , ski3 ) .
furthermore in sn isotopes ( @xmath3=100@xmath34132 ) , it is confirmed that the coefficients are suitable and the @xmath7 expansion is available .
the absolute values of @xmath31 are lager than those of @xmath4 , thus the analysis assuming @xmath31=@xmath122 is not effective , which are often used in previous study for non - relativistic interaction .
the @xmath6 estimation eq.([kci ] ) works well .
the isospin term @xmath8 in the finite system is smaller than the @xmath10 .
the magicity in a present work does not affect the conclusion . and
the pairing effect in sn isotopes are not so large , although the deformation due to the lack of pairing causes the coupling quadrupole and monopole modes , and disturbs the position of gmr .
we narrow down the values from the comparison experimental data with our results in a whole mass region , which indicates the range of @xmath8 is -305@xmath910 mev .
the incompressibility parameters corresponding to the range have @xmath4=225@xmath911 and @xmath11=-138@xmath918 mev , which is consistent with the previous study of slop parameter @xmath21 @xcite . as an indication of eq.([kti ] ) , the @xmath21 has an important role to decide @xmath11 . thus the studies related to @xmath21 should be also pressed forward in parallel .
the ordinary mass scaling of gmr has underestimation in heavy mass system , which may indicate the needs to reconsider the expression of incompressibility . to determine the isospin term @xmath8 , the systematic isgmr data in isobar nuclide are very effective
. they should be heavy system over @xmath3=100 at least , because in light system the nuclear speciality is showing up . when the heavy nucleus is used to analyse gmr
, we should note the deformation of open shell nuclei . if the monopole and other modes in deformed nucleus can be separated , the analysis data increases and the relation eos parameter and finite system will be more robust .
[ sec : con ]
the author greatly appreciates professor m. kimura for discussions and encouragements .
computational resources were mostly provided by the high performance computing system at research center for nuclear physics , osaka university .
h. sagawa , s. yoshida , g .- m . zeng , j .- z .
gu , and x .- z .
zhang , phys .
c*76 * , 034327 ( 2007 ) .
j. r. stone and p. -g .
reinhard , prog .
. phys . * 58 * , 587 ( 2007 ) .
j. piekarewicz and m. centelles , phys .
c*79 * , 054311 ( 2009 ) .
l. w. chen , c. m. ko , b .- a .
li , and j. xu , phys .
c * 82 * , 024321 ( 2010 ) .
m. dutra , o. loureno , j. s. s martins , a. delfino , j. r. stone and p. d. stevenson , phys .
c*85 * , 035201 ( 2012 ) .
b. a. brown , phys . rev
. lett . * 111 * , 232502 ( 2013 ) .
chen and j. piekarewicz , phys .
c*90 * , 044305 ( 2014 ) . m. dutra , o. loureno , s. s. avancini , b. v. carlson , a. delfino , d. p. menezes , c. providncia , s. typel and j. r. stone , phys .
c*90 * , 055203 ( 2014 ) .
t. li , et al . ,
lett . * 99 * , 162503 ( 2007 ) .
d. patel , et al .
b*718 * , 447 ( 2012 ) .
d. patel , et al .
b*726 * , 178 ( 2013 ) .
a. tamii et al .
* 107 * , 062502 ( 2011 ) .
m. warda , x. vias , x. roca - maza , and m. centelles , phys .
c*80 * , 024316 ( 2009 ) .
a. carbone , g. col , a. bracco , li - gang cao , p .
f. bortignon , f. camera , and o. wieland , phys .
c*81 * , 041301 ( r ) ( 2010 ) .
x. roca - maza , m. centelles , x. vi@xmath123as , and m. warda , phys .
* 106 * 252501 ( 2011 ) .
reinhard and w. nazarewicz , phys .
c*81 * , 051303 ( 2010 ) .
l. trippa , g. col , and e. vigezzi , phys .
c*77 * , 061304 ( r ) ( 2008 ) .
m. warda , m. centelles , x. vias , and x. roca - maza , phys .
c*89 * , 064302 ( 2014 ) .
s. ebata , t. nakatsukasa , t. inakura , k. yoshida , y. hashimoto and k. yabana , phys . rev .
c*82 * , 034306 ( 2010 ) .
s. ebata , t. nakatsukasa and t. inakura , phys .
c*90 * , 024303 ( 2014 ) .
g. scamps and d. lacroix , phys .
c*88 * , 044310 ( 2013 ) .
s. ebata and m. kimura , phys .
c*91 * , 014309 ( 2015 ) .
n. van giai and h. sagawa , phys .
b*106 * , 379 ( 1981 ) .
j. bartel , p. quentin , m. brack , c. guet , and h .- b .
hakansson , nucl .
a*386 * , 79 ( 1982 ) .
e. chabanat , p. bonche , p. haensel , j. meyer , and r. schaeffer , nucl .
a*635 * , 231 ( 1998 ) .
f. tondeur , m. brack , m. farine , and j. m. pearson , nucl .
a*420 * , 297 ( 1984 ) .
reinhard and h. flocard , nucl .
a*584 * , 467 ( 1995 ) .
e. khan , j. margueron , g. col , k. hagino and h. sagawa , phys .
c*82 * , 024322 ( 2010 ) .
p. avogadro and c. a. bertulani , phys .
c*88 * , 044319 ( 2013 ) .
j. piekarewicz , j. phys .
g : nucl . part .
* 37 * ( 2010 ) 064038 . | the isoscaler giant monopole resonances ( isgmr ) are computed using the canonical - basis time - dependent hartree - fock - bogoliubov theory ( cb - tdhfb ) with five kinds of skyrme parameter sets ( sgii , skm@xmath0 , sly4 , skt3 and ski3 ) . to extract the nuclear matter property from finite system , isgmrs of @xmath1=@xmath2 ( @xmath2=20 - 50 ) , isobar even - even nuclide for @xmath3=100 , 132 and sn
isotopes are analysed systematically .
the magnitude relation of nuclear incompressibility - parameter ( @xmath4 ) among skyrme parameter sets , can be corresponded to the peak positions of gmr in spherical isotopes over @xmath3=80 .
the parameters ( @xmath5 and @xmath6 ) which appear in expansion of the finite nucleus incompressibility @xmath7 , are determined for each skyrme parameter . from the comparison experimental data whole mass region and the present results ,
they indicate that the isospin dependent term @xmath8 is filtered as -305@xmath910 mev .
the incompressibility parameters of _ infinite _ system corresponding to our results are @xmath10=-340@xmath935 , @xmath4=225@xmath911 , and @xmath11=-138@xmath918 mev . |
media - based modulation ( mbm ) , a promising modulation scheme for wireless communications in multipath fading environments , is attracting recent research attention @xcite-@xcite .
the key features that make mbm different from conventional modulation are : @xmath3 mbm uses digitally controlled parasitic elements external to the transmit antenna that act as radio frequency ( rf ) mirrors to create different channel fade realizations which are used as the channel modulation alphabet , and @xmath4 it uses indexing of these rf mirrors to convey additional information bits .
the basic idea behind mbm can be explained as follows .
placing rf mirrors near a transmit antenna is equivalent to placing scatterers in the propagation environment close to the transmitter .
the radiation characteristics of each of these scatterers ( i.e. , rf mirrors ) can be changed by an on / off control signal applied to it .
an rf mirror reflects back the incident wave originating from the transmit antenna or passes the wave depending on whether it is off or on .
the on / off status of the mirrors is called as the ` mirror activation pattern ( map ) ' .
the positions of the on mirrors and off mirrors change from one map to the other , i.e. , the propagation environment close to the transmitter changes from one map to the other map .
note that in a rich scattering environment , a small perturbation in the propagation environment will be augmented by many random reflections resulting in an independent channel .
the rf mirrors create such perturbations by acting as controlled scatterers , which , in turn , create independent fade realizations for different maps . if @xmath0 is the number of rf mirrors used , then @xmath5 maps are possible .
if the transmitted signal is received through @xmath1 receive antennas , then the collection of @xmath5 @xmath1-length complex channel gain vectors form the mbm channel alphabet
. this channel alphabet can convey @xmath0 information bits through map indexing .
if the antenna transmits a symbol from a conventional modulation alphabet denoted by @xmath6 , then the spectral efficiency of mbm is @xmath7 bits per channel use ( bpcu ) .
an implementation of a mbm system consisting of 14 rf mirrors placed in a compact cylindrical structure with a dipole transmit antenna element placed at the center of the cylindrical structure has been reported in @xcite .
early reporting of the idea of using parasitic elements for index modulation purposes ( in the name ` aerial modulation ' ) can be found in @xcite,@xcite .
mbm has been shown to possess attractive performance attributes , particularly when the number of receive antennas is large @xcite-@xcite .
specifically , mbm with @xmath0 rf mirrors and @xmath1 receive antennas over a multipath channel has been shown to asymptotically ( as @xmath2 ) achieve the capacity of @xmath1 parallel awgn channels @xcite .
this suggests that mbm can be attractive for use in massive mimo systems which typically employ a large number of receive antennas at the bs .
however , the literature on mbm so far has focused mainly on single - user ( point - to - point ) communication settings .
our first contribution in this paper is that , we report mbm in multiuser massive mimo settings and demonstrate significant performance advantages of mbm compared to conventional modulation . for example , a bit error performance achieved using 500 receive antennas at the bs in a massive mimo system using conventional modulation can be achieved using just 128 antennas with multiuser mbm .
even multiuser spatial modulation ( sm ) and generalized spatial modulation ( gsm ) @xcite-@xcite in the same massive mimo settings require more than 200 antennas to achieve the same bit error performance .
this suggests that multiuser mbm can be an attractive scheme for use in the uplink of massive mimo systems .
the second contribution relates to exploitation of the inherent sparsity in multiuser mbm signal vectors for low - complexity signal detection at the bs receiver .
we resort to compressive sensing ( cs ) based sparse recovery algorithms for this purpose .
several efficient sparse recovery algorithms are known in the literature@xcite-@xcite .
we propose a multiuser mbm signal detection scheme that employs greedy sparse recovery algorithms like orthogonal matching pursuit ( omp)@xcite , compressive sampling matching pursuit ( cosamp ) @xcite , and subspace pursuit ( sp)@xcite .
simulation results show that the proposed detection scheme using sp achieves very good performance ( e.g. , significantly better performance compared to mmse detection ) at low complexity .
this demonstrates that cs based sparse signal recovery approach is a natural and efficient approach for multiuser mbm signal detection in massive mimo systems .
the rest of the paper is organized as follows .
the multiuser mbm system model is introduced in sec .
the performance of multiuser mbm with maximum likelihood detection is presented in sec .
[ sec3 ] . the proposed sparsity - exploiting detection scheme for multiuser mbm signal detection and its performance in massive mimo systems
are presented in sec .
conclusions are presented in sec .
consider a massive mimo system with @xmath8 uplink users and a bs with @xmath1 receive antennas ( see fig .
[ mbm_mimo ] ) , where @xmath8 is in the tens ( e.g. , @xmath9 ) and @xmath1 is in the hundreds ( @xmath10 ) .
the users employ mbm for signal transmission .
each user has a single transmit antenna and @xmath0 rf mirrors placed near it . in a given channel use , each user selects one of the @xmath5 mirror activation patterns ( maps ) using @xmath0 information bits .
a mapping is done between the combinations of @xmath0 information bits and the maps .
an example mapping between information bits and maps is shown in table [ table1 ] for @xmath11 .
the mapping between the possible maps and information bits is made known a priori to both transmitter and receiver for encoding and decoding purposes , respectively .
.mapping between information bits and maps for @xmath11 . [ cols="^,^,^",options="header " , ] apart from the bits conveyed through the choice of a map in a given channel use as described above , a symbol from a modulation alphabet @xmath12 ( e.g. , qam , psk ) transmitted by the antenna conveys an additional @xmath13 bits .
therefore , the spectral efficiency of a @xmath8-user mbm system is given by @xmath14 for example , a multiuser mbm system with @xmath15 , @xmath11 , and 4-qam has a system spectral efficiency of 16 bpcu .
an important point to note here is that the spectral efficiency per user increases linearly with the number of rf mirrors used at each user . to introduce the multiuser mbm signal set and the corresponding received signal vector at the bs ,
let us first formally introduce the single - user mbm signal set .
the mbm channel alphabet of a single user is the set of all channel gain vectors corresponding to the various maps of that user .
let us define @xmath16 , where @xmath17 is the number of possible maps corresponding to @xmath0 rf mirrors .
let @xmath18 denote the @xmath19 channel gain vector corresponding to the @xmath20th map of the @xmath21th user , where @xmath22^t$ ] , @xmath23 is the channel gain corresponding to the @xmath20th map of the @xmath21th user to the @xmath24th receive antenna , @xmath25 , @xmath26 , and @xmath27 , and the @xmath23s are assumed to be i.i.d . and distributed as @xmath28 .
the mbm channel alphabet for the @xmath21th user , denoted by @xmath29 , is then the collection of these channel gain vectors , i.e. , @xmath30 .
the mbm channel alphabet of each user is estimated at the bs receiver through pilot transmission before data transmission .
the number of pilot channel uses needed for the estimation of each user s channel alphabet grows exponentially in @xmath0 .
it is also noted that , while the mbm channel alphabet of each user needs to be known at the bs receiver for detection purposes , the users transmitters need not know their channel alphabets . define @xmath31 .
the single - user mbm signal set , denoted by @xmath32 , is the set of @xmath33-sized mbm signal vectors given by @xmath34^t , s_q \in \mathbb{a } , \label{ss}\end{aligned}\ ] ] where @xmath20 is the index of the map .
that is , an mbm signal vector @xmath35 in ( [ ss ] ) means a complex symbol @xmath36 being transmitted on a channel with an associated channel gain vector @xmath37 , where @xmath37 is the @xmath38 channel gain vector corresponding to the @xmath20th map .
therefore , the @xmath38 received signal vector corresponding to a transmitted mbm signal vector @xmath35 can be written as @xmath39 where @xmath40 is the awgn noise vector with @xmath41 .
the size of the single - user mbm signal set is @xmath42 .
for example , if @xmath11 and @xmath43 ( i.e. , bpsk ) , then @xmath44 , and the corresponding mbm signal set is given by @xmath45 with the above definitions of single - user mbm channel alphabet and signal set , the multiuser mbm signal set with @xmath8 users is given by @xmath46 .
let @xmath47 denote the transmit mbm signal vector from the @xmath21th user .
let @xmath48^t \in \mathbb{s}_{{\tiny \mbox{mu - mbm}}}$ ] denote the vector comprising of the transmit mbm signal vectors from all the @xmath8 users .
let @xmath49 denote the channel gain matrix given by @xmath50 $ ] , where @xmath51 \in \mathbb{c}^{n_r \times m}$ ] , and @xmath18 is the channel gain vector of the @xmath21th user corresponding to @xmath20th map as defined before .
the @xmath38 multiuser received signal vector at the bs is then given by @xmath52 where @xmath53 is the @xmath38 awgn noise vector with @xmath54 .
in this section , we analyze the ber performance of multiuser mbm under maximum likelihood ( ml ) detection .
we obtain an upper bound on the ber which is tight at moderate to high snrs .
we also present a comparison between the ber performance of multiuser mbm and those of other multiuser schemes that employ conventional modulation , spatial modulation , and generalized spatial modulation .
the ml detection rule for the multiuser mbm system model in ( [ sys ] ) is given by @xmath55 which can be written as @xmath56 the pairwise error probability ( pep ) that the receiver decides in favor of the signal vector @xmath57 when @xmath58 was transmitted , given the channel matrix @xmath59 can be written as @xmath60 defining @xmath61 , we observe that @xmath62 .
therefore , we can write @xmath63 where @xmath64 . the conditional pep expression in can be written as @xmath65 where @xmath66 and @xmath67 are @xmath68th entries of @xmath58 and @xmath57 , respectively , and @xmath69 is the @xmath68th column of @xmath59 .
the argument of @xmath70 in has the central @xmath71-distribution with @xmath72 degrees of freedom .
the computation of the unconditional peps requires the expectation of @xmath70 with respect to @xmath59 , which can be obtained as follows@xcite : @xmath73 \nonumber \\ & = f(\alpha)^{n_r } \sum_{i=0}^{n_r-1 } \binom{n_r-1+i}{i } ( 1-f(\alpha))^i , \end{aligned}\ ] ] where @xmath74 , @xmath75 , and @xmath76 .
now , an upper bound on the bit error probability using union bound can be obtained as @xmath77 where @xmath78 is the hamming distance between the bit mappings corresponding to @xmath58 and @xmath57 .
we evaluated the ber performance of multiuser mbm ( mu - mbm ) using the ber upper bound derived above as well as simulations .
for the purpose of initial comparisons with other systems , we consider a mu - mbm system with @xmath79 , @xmath80 , @xmath81 , bpsk , and 4 bpcu per user .
let @xmath82 and @xmath83 denote the number transmit antennas and transmit rf chains , respectively , at each user .
note that in the considered mu - mbm system , each user uses one transmit antenna and one transmit rf chain , i.e. , @xmath84 .
we compare the performance of the above mu - mbm system with those of three other multiuser systems which use @xmath3 conventional modulation ( cm ) , @xmath4 spatial modulation ( sm ) , and @xmath85 generalized spatial modulation ( gsm ) .
the multiuser system with conventional modulation ( mu - cm ) uses @xmath84 at each user and employs 16-qam to achieve the same spectral efficiency of 4 bpcu per user .
the multiuser system with sm ( mu - sm ) uses @xmath86 , @xmath87 , and 8-qam , achieving a spectral efficiency of @xmath88 bpcu per user .
the multiuser system with gsm ( mu - gsm ) uses @xmath89 , @xmath90 , and bpsk , achieving a spectral efficiency of @xmath91 bpcu per user .
figure [ ml ] shows the ber performance of the mu - mbm , mu - cm , mu - sm , and mu - gsm systems described above .
first , it can be observed that the analytical upper bound is very tight at moderate to high snrs .
next , in terms of performance comparison between the considered systems , the following inferences can be drawn from fig .
[ ml ] . *
the mu - mbm system achieves the best performance among all the four systems considered .
for example , mu - mbm performs better by about 5 db , 4 db , 2.5 db compared to mu - cm , mu - sm , and mu - gsm systems , respectively , at a ber of @xmath92 . *
the better performance of mu - mbm can be attributed to more bits being conveyed through mirror indexing , which allows mu - mbm to use lower - order modulation alphabets ( bpsk ) compared to other systems which may need higher - order alphabets ( 8-qam , 16-qam ) to achieve the same spectral efficiency . * mu - mbm performs better than mu - gsm though both use bpsk in this example .
this can be attributed to the good distance properties of the mbm signal set @xcite . , @xmath80 , 4 bpcu per user , and ml detection .
analysis and simulations.,width=340,height=236 ] note that though the results in fig .
[ ml ] illustrate the performance superiority of mu - mbm over mu - cm , mu - sm , and mu - gsm , they are presented only for a small system with @xmath79 and @xmath80 .
this is because ml detection is prohibitively complex for systems with large @xmath8 and @xmath1 ( ml detection is exponentially complex in @xmath8 ) .
however , massive mimo systems are characterized by @xmath8 in the tens and @xmath1 in the hundreds . therefore , low - complexity detection schemes which scale well for such large - scale mu - mbm systems are needed . to address this need , we resort to exploiting the inherent sparse nature of the mbm signal vectors , and devise a compressive sensing based detection algorithm in the following section .
it is evident from the example signal set in that the mbm signal vectors are inherently sparse .
an mbm signal vector has only one non - zero element out of @xmath17 elements , leading to a sparsity factor of @xmath93 .
for example , consider an mbm signal set with @xmath94 and @xmath95 . out of 16 elements in a signal vector ,
only one element is non - zero resulting in a sparsity factor of @xmath96 .
exploitation of this inherent sparsity to devise detection algorithms can lead to efficient signal detection at low complexities .
accordingly , we propose a low - complexity mu - mbm signal detection scheme that employs compressive sensing based sparse reconstruction algorithms like omp , cosamp , and sp . we first model the mu - mbm signal detection problem as a sparse reconstruction problem and then employ greedy algorithms for signal detection .
sparse reconstruction is concerned with finding an approximate solution to the following problem @xcite : @xmath97 where @xmath98 is called the measurement matrix , @xmath99 is the complex input signal vector , @xmath100 is the noisy observation corresponding to the input signal , and @xmath101 is the complex noise vector . the mu - mbm signal detection problem at the bs in ( [ sys ] ) can be modeled as a sparse recovery problem in , with the measurement matrix being the channel matrix @xmath49 , the noisy observation being the received signal vector @xmath102 , and the input being the mu - mbm transmit signal vector @xmath103 .
the noise vector is additive complex gaussian with @xmath104 .
greedy algorithms achieve sparse reconstruction in an iterative manner .
they decompose the problem of sparse recovery into a two step process ; recover the support of the sparse vector first , and then obtain the non - zero values over this support .
for example , omp starts with an initial empty support set , an initial solution @xmath105 , and an initial residue @xmath106 . in each step ,
omp updates one coordinate of the vector @xmath107 based on the correlation values between the residue vector and the columns of the @xmath108 matrix . in the @xmath21th iteration , an element @xmath109 given by @xmath110
is added to the support set , where @xmath111 is the @xmath112th column of @xmath113 , and @xmath114 and @xmath115 are the support set and residue after @xmath116 iterations , respectively .
the entries of @xmath107 corresponding to the obtained support set are computed using least squares .
this process is iterated till the stopping criteria is met .
the stopping criteria can be either a specified error threshold or a specified level of sparsity . in the sp algorithm , instead of updating one coordinate of @xmath107 at a time as in omp
, @xmath8 coordinates are updated at once .
the major difference between omp and sp is the following . in omp ,
the support set is generated sequentially .
it starts with an empty set and adds one element in every iteration to the existing support set .
an element added to the support set can not be removed until the algorithm terminates .
in contrast , sp provides flexibility of refining the support set in every iteration .
cosamp is similar to sp except that it updates @xmath117 coordinates in each iteration to the support set instead of updating @xmath8 coordinates as in sp .
cosamp and sp have superior reconstruction capability comparable to convex relaxation methods @xcite,@xcite .
* algorithm 1 * shows the listing of the pseudo - code of the proposed sparsity - exploiting detection algorithm for mu - mbm signals .
inputs : @xmath118 initialize : @xmath119 * repeat * @xmath120 @xmath121 * if * @xmath122 * for * @xmath123 @xmath124 * end for * * break * ; * else * @xmath125 * end if * * until * @xmath126 output : the estimated mu - mbm signal vector @xmath127^t\ ] ] [ alg1 ] sr in * algorithm 1 * denotes the sparse recovery algorithm , which can be any one of omp , cosamp , and sp . the signal vector reconstructed by the sparse recovery algorithm is denoted by @xmath128 . detecting the mu - mbm signal vector
involves detecting the mbm signal vector transmitted by each user .
an mbm signal vector from a user has exactly one non - zero entry out of @xmath17 entries as observed in the example mbm signal set in .
hence , sr is expected to reconstruct a mu - mbm signal vector such that the mbm signal sub - vector corresponding to a given user has only one non - zero entry . but this constraint on the expected support set is not built in the general sparse recovery algorithms . in general , a sparse recovery algorithm can output @xmath8 non - zero elements at any of the @xmath129 locations of @xmath128 . to overcome this issue
, we define user activity pattern ( uap ) , denoted by @xmath130 , as a @xmath8-length vector with @xmath21th entry as @xmath131 if there is at least one non - zero entry in the @xmath21th user s recovered mbm signal vector , and @xmath132 otherwise .
a valid reconstructed signal vector is one which has all ones in @xmath130 .
sr is used multiple times with a range of sparsity estimates starting from @xmath8 ( @xmath133 in the algorithm listing ) till the valid uap is obtained ( i.e. , till the algorithm reconstructs at least one non - zero entry for each user s mbm signal vector ) . in the algorithm listing , @xmath134 denotes the uap at the @xmath112th iteration . on recovering an @xmath128 with valid uap
, the mbm signal vector of each user is mapped to the nearest ( in the euclidean sense ) mbm signal vector in @xmath32 .
this is shown in the step 8 in the algorithm listing , where @xmath135 denotes the recovered mbm signal vector of the @xmath21th user and @xmath136 denotes the mbm signal vector to which @xmath135 gets mapped to .
finally , the mu - mbm signal vector is obtained by concatenating the detected mbm signal vectors of all the users , i.e. , , @xmath137^t$ ] .
the decoding of information bits from the detected mbm signal vector of a given user involves decoding of mirror index bits and qam symbol bits of that user .
the mirror index bits are decoded from the map of the detected mbm signal vector and the qam bits are decoded from the detected qam symbol . in this subsection
, we present the ber performance of mu - mbm systems in a massive mimo setting ( i.e. , @xmath8 in the tens and @xmath1 in the hundreds ) when the proposed * algorithm 1 * is used for mu - mbm signal detection at the bs . in the same massive mimo setting
, we evaluate the performance of other systems that use conventional modulation ( mu - cm ) , spatial modulation ( mu - sm ) , and generalized spatial modulation ( mu - gsm ) , and compare them with the performance achieved by mu - mbm . the proposed * algorithm 1 *
is also used for the detection of mu - sm and mu - gsm .
it is noted that the mu - sm and mu - gsm signal vectors are also sparse to some extent ; the sparsity factors in mu - sm and mu - gsm are @xmath138 and @xmath139 , respectively .
so the use of the proposed algorithm for detection of these signals is also appropriate .
ml detection is used to detect mu - cm signals ( this is possible for mu - cm with sphere decoding for @xmath140 , i.e. , 32 real dimensions ) .
_ mu - mbm performance using proposed algorithm : _ figure [ fig : sparsealgos ] shows the performance of mu - mbm system using the proposed algorithm with @xmath3 omp , @xmath4 cosamp , and @xmath85 sp . mmse detection performance is also shown for comparison .
a massive mimo system with @xmath140 and @xmath141 is considered .
each user uses @xmath142 , @xmath87 , @xmath143 , and 4-qam .
this results in a spectral efficiency of 8 bpcu per user , and a sparsity factor of @xmath144 . from fig .
[ fig : sparsealgos ] , we observe that the proposed algorithm with omp , cosamp , and sp achieve significantly better performance compared to mmse . among
the the use of omp , cosamp , and sp in the proposed algorithm , use of sp gives the best performance .
this illustrates the superior reconstruction / detection advantage of the proposed algorithm with sp .
we will use the proposed algorithm with sp in the subsequent performance results figures .
it is noted that the complexity of proposed algorithm is also quite favorable ; the complexity of the proposed algorithm with sp and that of mmse are @xmath145 and @xmath146 , respectively . , @xmath141 , @xmath142 , @xmath87 , @xmath143 , 4-qam , 8bpcu per user , using the proposed detection algorithm .
mmse detection performance is also shown for comparison.,width=340,height=236 ] _ performance of mu - mbm , mu - sm , mu - gsm : _ figure [ fig : spdetection ] shows a ber performance comparison between mu - mbm , mu - cm , mu - sm , and mu - gsm in a massive mimo setting with @xmath140 and @xmath141 .
the proposed algorithm with sp is used for detection in mu - mbm , mu - sm , and mu - gsm .
ml detection is used for mu - cm .
the spectral efficiency is fixed at 5 bpcu per user for all the four schemes .
mu - mbm achieves this spectral efficiency with @xmath142 , @xmath87 , @xmath81 , and 4-qam .
mu - cm uses @xmath142 , @xmath87 , and 32-qam to achieve 5 bpcu per user . to achieve the same 5 bpcu per user ,
mu - sm uses @xmath89 , @xmath87 , and 8-qam , and mu - gsm uses @xmath147 , @xmath90 , and bpsk .
the sparsity factors in mu - mbm , mu - sm , and mu - gsm are @xmath148 , @xmath149 , and @xmath150 , respectively .
it can be seen that , mu - mbm clearly outperforms mu - cm , mu - sm , and mu - gsm .
for example , at a ber of @xmath92 , mu - mbm outperforms mu - cm , mu - gsm , and mu - sm by about 7 db , 5 db , and 4 db , respectively .
the performance advantage of mu - mbm can be mainly attributed to its better signal distance properties @xcite .
mu - mbm is also benefited by its lower sparsity factor as well as the possibility of using lower - order qam size because of additional bits being conveyed through indexing mirrors . ,
@xmath141 , and 5 bpcu per user.,width=340,height=236 ] [ fig : spdetection ] _ effect of number of bs receive antennas _ :
figure [ fig : effectofnr ] shows an interesting result which demonstrates mu - mbm s increasing performance gain compared to mu - cm , mu - sm , and mu - gsm as the number of bs receive antennas is increased .
a massive mimo system with @xmath140 and 5 bpcu per user is considered .
the parameters of the four schemes are the same as those in fig .
[ fig : spdetection ] except that here snr is fixed at 4 db and @xmath1 is varied from 48 to 624 .
it is interesting to observe that a performance that could be achieved using 500 antennas at the bs in a massive mimo system that uses conventional modulation ( @xmath151 ber for mu - cm at @xmath152 with ml detection ) can be achieved using just 128 antennas when mu - mbm is used ( @xmath151 ber for mu - mbm at @xmath141 with proposed detection ) .
mu - sm and mu - gsm also achieve better performance compared to mu - cm , but they too require more than 200 antennas to achieve the same ber . this increasing performance advantage of mu - mbm for increasing @xmath1 can be mainly attributed to its better signal distance properties particularly when @xmath1 is large @xcite .
this indicates that multiuser mbm can be a very good scheme for use in the uplink of massive mimo systems . in a massive mimo setting with @xmath140 , 5 bpcu per user , and snr = 4 db.,width=340,height=236 ]
we investigated the use of media - based modulation ( mbm ) , a recent and attractive modulation scheme that employs rf mirrors ( parasitic elements ) to convey additional information bits through indexing of these mirrors , in massive mimo systems .
our results demonstrated significant performance advantages possible in multiuser mbm compared to multiuser schemes that employ conventional modulation , spatial modulation , and generalized spatial modulation .
motivated by the possibility of exploiting the inherent sparsity in multiuser mbm signal vectors , we proposed a detection scheme based on compressive sensing algorithms like omp , cosamp , and subspace pursuit .
the proposed detection scheme was shown to achieve very good performance ( e.g. , significantly better performance compared to mmse detection ) at low complexity , making it suited for multiuser mbm signal detection in massive mimo systems .
channel estimation , effect of imperfect knowledge of the channel alphabet at the receiver , and effect of spatial correlation are interesting topics for further investigation .
o. n. alrabadi , a. kalis , c. b. papadias , and r. prasad , `` a universal encoding scheme for mimo transmission using a single active element for psk modulation schemes , '' _ ieee trans .
wireless commun .
5133 - 5142 , oct . 2009 . m.
di renzo , h. haas , a. ghrayeb , s. sugiura , and l. hanzo , `` spatial modulation for generalized mimo : challenges , opportunities and implementation , '' _ proc . of the ieee _ , vol .
56 - 103 , jan . 2014 .
j. wang , s. jia , and j. song , `` generalised spatial modulation system with multiple active transmit antennas and low complexity detection scheme , '' _ ieee trans .
wireless commun_. , vol .
1605 - 1615 , apr .
2012 .
d. donoho , y. tsaig , i. drori , and j .- l .
starck , `` sparse solution of underdetermined systems of linear equations by stagewise orthogonal matching pursuit , '' _ ieee trans .
inform . theory _
2 , pp . 1094 - 1121 , feb .
2012 .
m. s. alouini and g. goldsmith , `` a unified approach for calculating error rates of of linearly modulated signals over generalized fading channels , '' _ ieee trans .
9 , pp . 1324 - 1334 , sep . | in this paper , we consider _ media - based modulation ( mbm ) _ , an attractive modulation scheme which is getting increased research attention recently , for the uplink of a massive mimo system .
each user is equipped with one transmit antenna with multiple radio frequency ( rf ) mirrors ( parasitic elements ) placed near it . the base station ( bs )
is equipped with tens to hundreds of receive antennas .
mbm with @xmath0 rf mirrors and @xmath1 receive antennas over a multipath channel has been shown to asymptotically ( as @xmath2 ) achieve the capacity of @xmath1 parallel awgn channels .
this suggests that mbm can be attractive for use in massive mimo systems which typically employ a large number of receive antennas at the bs . in this paper
, we investigate the potential performance advantage of multiuser mbm ( mu - mbm ) in a massive mimo setting .
our results show that multiuser mbm ( mu - mbm ) can significantly outperform other modulation schemes .
for example , a bit error performance achieved using 500 receive antennas at the bs in a massive mimo system using conventional modulation can be achieved using just 128 antennas using mu - mbm .
even multiuser spatial modulation , and generalized spatial modulation in the same massive mimo settings require more than 200 antennas to achieve the same bit error performance .
also , recognizing that the mu - mbm signal vectors are inherently sparse , we propose an efficient mu - mbm signal detection scheme that uses compressive sensing based reconstruction algorithms like orthogonal matching pursuit ( omp ) , compressive sampling matching pursuit ( cosamp ) , and subspace pursuit ( sp ) . |
the concept of electrostatic screening has been well known since the work by debye and hckel of early @xmath1 century @xcite . in recent years , screening by strongly charged ions was found to result in counterintuitive phenomena such as attraction between like - charged macroions @xcite , and inversion of macroion charges .
charge inversion was studied by experiments using both colloids and biological materials , by analytical theories @xcite , and by monte carlo and molecular dynamics ( md ) simulations @xcite .
experimentally , the most direct method to observe charged colloids and macroions is electrophoresis .
this method is also the prime candidate for the technique of observing charge inversion .
although straightforward , this approach involves many questions upon a closer look .
for example , does the macroion drift along with its adsorbed multivalent ions when an external electric field is applied ? how many multivalent ions are attached to the macroion strongly enough to drift together ?
how is the drift affected by the solvent viscosity and the counterflow of monovalent ions ?
what happens when the field becomes very strong ?
what is the field strength sufficient to disrupt the charged complex ?
these are the fundamental questions necessary to address in order to bridge theoretical concepts of charge inversion and experimental observations .
it should be born in mind that electrophoresis in general has quite a few delicate aspects .
some peculiar ones attracted significant attentions recently @xcite ( see also the review article @xcite and references therein ) .
the electric field acts not only on the macroions , but on every ion in a solution . in many cases , this leads to effective screening of hydrodynamic interactions which otherwise may be very significant . in the simulations reported below , we have actually observed such short - range screening of hydrodynamic interactions in the system comprising of a macroion , counterions and coions ( sec.[sec.3.2.1 ] ) .
one of the difficulties in simulating charge inversion under electrophoresis consists in subtle interactions of a macroion with surrounding ions and neutral solvent . a naive use of the langevin equation , assuming that every ion ( radius @xmath2 ) in the system is subject to stokesian friction @xmath3 and white noise random forces that balance the friction through the fluctuation - dissipation theorem , is not justifiable .
a simple counterexample would be two closely located spheres .
since other particles ( either ions or neutral solvent ) are excluded from the volume between spheres , neither their corresponding friction forces nor random forces add to each other .
one way of going around this problem is to incorporate macroion - solvent interactions using the oseen tensor @xcite .
this is , however , not easy to implement in numerical simulations , because the interactions produce complicated spatial correlations among random forces .
therefore , we address this problem by a direct approach introducing explicit neutral particles to deal with the macroion - salt - solvent interactions in the molecular dynamics simulations .
the explicit simulation of the solvent molecules is of course costly . in this paper , with the limited computational resources , we restrict ourselves to the system with only the @xmath0 salt , and no @xmath4 salt .
it is needless to say that in real water solvent there is always some amount of @xmath5 salt . in this sense ,
our present paper demonstrates the principle that charge inversion is a phenomenon observable by a direct electrophoresis experiment .
further study including the @xmath4 salt will be required to compare results with realistic systems . here
, we will specify the deviations arising from the lack of the @xmath4 salt .
also , we will confine ourselves to the study of a single macroion interacting with surrounding salt ions and solvent .
our plan in this study is to examine electrophoresis of a spherically shaped , uniformly charged macroion .
we will systematically measure the mobility of the drifting macroion complex placed under an external electric field by molecular dynamics simulations .
we first show that charge inversion does take place in a solution containing multivalent counterions , as manifested by the inverse mobility under the weak electric field .
we then look at the dependences of the reversed mobility on the parameters , such as the concentration of co- and counterions and the surface charge density of the macroion .
finally , we consider the strong electric field regime and show that the strong field disrupts the charge - inverted macroion complex and terminates the charge inversion phenomenon .
we adopt the following model , with @xmath6 , @xmath7 , and @xmath8 being the units of length , charge and mass , respectively .
( we have in mind @xmath9 and @xmath10 a.m.u . ) a macroion with negative charge @xmath11 between @xmath12 and @xmath13 is surrounded by @xmath14 counterions of a positive charge @xmath15 and @xmath16 coions of a negative charge @xmath17 .
the system is maintained in overall charge neutrality , @xmath18 , which determines @xmath14 for a given @xmath19 .
the mass of the macroion is @xmath20 , and the mass of the co- and counter - ions is @xmath8 .
we also include @xmath21 neutral particles with mass @xmath22 , where we note the mass of water molecule against that of k@xmath23 or ca@xmath24 ions .
approximately one neutral particle is located in every volume element @xmath25@xmath26 inside the simulation domain , excluding the locations already occupied by the macroion and other ions , which typically yields 8000 neutral particles .
these particles are confined in a rectangular box of size @xmath27 , with periodic boundary conditions in all three directions .
most of the runs are performed for @xmath28 , except one series of the runs intended to test the finite size effect of the domain ( fig.[fig.mu_l ] ) described in sec.[sec.3.2.1 ] .
in addition to the coulomb forces , all particles interact through the repulsive lennard - jones potential @xmath29 $ ] for @xmath30 otherwise . here
@xmath31 is the position vector of the @xmath32-th particle , and @xmath33 is the sum of the radii of two interacting particles , which are chosen as follows : radius of the macroion , @xmath34 , is between @xmath35 and @xmath36 , counterions and coions have radius @xmath6 , and neutral particles @xmath37 .
we relate @xmath38 with the temperature by @xmath39 , and choose @xmath40 ( we assume spatially homogeneous dielectric constant @xmath41 ) .
the bjerrum length is thus @xmath42 . for the parameters of this temperature ,
the valence @xmath43 , and the number of coions @xmath44 , the ionic strength becomes @xmath45 . after knowing the ionic strength , one is tempted to compute the debye length which turns out to be @xmath46 .
we should emphasize that this number does not have much meaning for the system under study , because we work in the domain very far from applicability of the standard debye - hckel theory .
in particular , the average number of ions in the volume @xmath47 turns out smaller than unity .
this is by no means surprising , because there is no charge inversion in the debye - hckel theory , and to examine charge inversion we need to go to the regime where this theory fails .
since we do not have the @xmath4 salt in our simulation , we should keep in mind that correlations between strongly charged @xmath48-ions may be significant even away from the macroion .
indeed , in the theory of charge inversion @xcite , the role of correlations is emphasized for the @xmath48-ions in the vicinity of the macroion , where their concentration is particularly large . in our system , the concentration of the @xmath48-ions is not very small even in the bulk , and we face the situations in which correlations between the @xmath48-ions away from the macroion affect our results . for such cases , we make additional runs with the reduced concentration in the @xmath0 salt .
however , we do not use this reduced concentration for all the runs , because such system is more prone to noises and fluctuations , requiring larger statistics to obtain reliable results .
calculation of the coulomb forces under the periodic boundary conditions involves the charge sum in the first brillouin zone and their infinite mirror images ( the ewald sum@xcite ) .
the sum is calculated with the use of the pppm algorithm @xcite .
we use @xmath49 spatial meshes for the calculation of the reciprocal space contributions to the coulomb forces , with the ewald parameter @xmath50 and the real - space cutoff @xmath51 , where @xmath52 is the radius of the @xmath32-th ion .
a uniform electric field @xmath53 is applied in the @xmath54-direction .
when starting the molecular dynamics simulation run , we prepare an initial state by randomly positioning all the ions and neutral particles in the simulation domain and giving maxwell - distributed random velocities corresponding to the temperature @xmath55 .
we integrate the newton equations of motion with the use of the leapfrog method @xcite , which is equivalent to the verlet algorithm . in the absence of the electric field ( @xmath56 )
, our system is closed , and its energy is conserved . after an initial transient phase , the distribution of velocities relaxes to a maxwellian , corresponding to an equilibrium sampling of the microcanonical ensemble .
this new maxwell distribution has the temperature @xmath57 , which is a little higher than @xmath55 , because of the release of the potential energy due to screening , i.e. , local balancing of charges .
we adjust @xmath55 such that @xmath58 .
this makes @xmath38 to be the unique relevant scale of energy , and , accordingly , we put @xmath59 as the unit of time .
we choose @xmath60 as the integration time step .
the simulation runs are executed up to @xmath61 .
when an external electric field is present , it performs work on the system .
the corresponding energy , which is the joule heat , is transferred to background neutral particles through collisions with accelerated ions . in our work
, we simulate an electrophoretic bath that is kept at a constant temperature @xmath57 . for this purpose , we pretend that all neutral particles go through the thermal bath of infinite heat capacitance , whenever they cross the boundaries of the simulation domain ( at the center of which the macroion is located at every moment ) .
operationally , we refresh the velocities of the neutral particles according to the thermal distribution when they cross the domain boundaries . this procedure maintains temperature stably to within 5% .
two closely related factors are potentially dangerous as they might affect the molecular dynamics simulation results .
one factor is the finite - size of the simulation domain , and the other is the long - range character of both hydrodynamic and coulomb interactions .
these problems become particularly important because some of the methods to simulate a constant temperature thermal bath are believed to lead to the screening of hydrodynamic interactions .
this is obviously not acceptable in the simulation where the long range character of hydrodynamic interactions is expected to be important @xcite .
following @xcite , we argue that hydrodynamic interactions are in fact effectively screened in our system and , therefore , that the domain size in our simulation is quite sufficient and the heat bath procedure described above is benign and reliable .
to understand the situation , it is worth discussing the major point - the screening of hydrodynamic interactions .
to begin with , why are hydrodynamic interactions long ranged ?
that fact can be understood well from the point of view of momentum conservation .
consider a particle immersed in a fluid and suppose that we pull this particle with an externally applied force ( such as gravity ) .
obviously , this force transfers momentum into the system and , however large the container may be , this momentum must be transported away through the container walls .
this necessitates the long range character of the hydrodynamic field .
more accurately , if we surround our object by an arbitrarily large closed surface , then ( under the stationary conditions ) the outflow of momentum through this surface must be equal to the inflow of momentum due to the external force .
because of the obvious analogy with the gauss theorem in electrostatics , we see that hydrodynamic field is just as long ranged as the coulomb field ( even though it has a more complicated vector structure ) .
the above description must be modified significantly when the applied external force is due to the electric field and the solution is neutral as a whole . in this case , there is no overall inflow of momentum into the system , and therefore , there should not be any outflow through the walls of the container .
more specifically , if there is one negatively charged macroion as in our simulation , it is surrounded by positively charged counterions and negatively charged coions such that the total charge of the crowd effectively vanishes at some finite distance . in the simple case of the debye - hckel theory
, this happens at about the debye screening length @xmath62 from the macroion surface .
therefore , no momentum is transported further away , and hydrodynamic interactions are screened at the distances of the order of @xmath62 @xcite . in this paper
, we treat a more complicated situation in which the debye - hckel theory does not apply and it is not easy to judge _ a priori _ at which distances the hydrodynamic interactions are screened . nevertheless ,
since the system is neutral , hydrodynamic interactions must be screened .
we therefore perform special test ( described in sec.[sec.3.2.1 ] ) looking at the dependence of the macroion drift on the simulation domain size .
we find that the drift is essentially size - independent at @xmath63 which is the direct manifestation of the screening of hydrodynamic interactions .
since hydrodynamic interactions are screened , our simulation is not very sensitive to the method of maintaining the constant temperature . to test it
, we have performed separate runs at weak electric fields , @xmath64 , in which case we can run the simulation even without any heat drainage for a significant period of time before any noticeable heating of the system ; the results of these control runs are within error bars of the data obtained using the thermal bath ( fig.[fig.mu_qsurf ] ) .
our simulation results are shown in figs.1 - 7 . fig .
[ fig.1 ] is a birds - eye view of ( a ) all the ions and ( b ) the vicinity of the macroion .
counterions are shown in light blue and coions in dark blue ( neutral atoms are not shown ) . in this figure ,
the macroion charge is taken to be @xmath65 , its radius @xmath66 , counterion valence @xmath67 , and the electric field @xmath68 .
it is seen that the macroion is predominantly covered by the counterions . as in the case without the electric field @xcite , the radially integrated charge has a sharp positive peak at a distance about @xmath6 from the macroion surface .
this peak is due to the positive counterions being adsorbed on the macroion surface .
the value of the peak charge under the conditions of fig.[fig.1 ] is @xmath69 demonstrates the time history of ( a ) the `` peak '' charge and ( b ) the macroion drift speed for the parameters of fig.[fig.1 ] . at time @xmath70 , we switch on the external electric field .
there is a short transient phase during which a charge - inverted complex is formed through adsorption of counterions to the macroion and condensation of coions on the counterions .
this process is reflected in a rather quick rise in @xmath71 , as is shown in fig.[fig.t_hist](a ) .
after the transient phase , we observe a drift of the macroion in the _ positive _ direction along the applied field .
the fact that the drift velocity is positive for the negative bare charge of the macroion ( @xmath72 ) is a direct manifestation of charge inversion such that counterions are so strongly bound that they pull the macroion with their motion .
note that the drift velocity shown in fig.[fig.t_hist](b ) is small compared to the thermal velocity @xmath73 of neutral particles , @xmath74 . under this condition ,
exchange of momentum between the macroion and neutral particles is slow , and it requires many collisions ( compare the similar system in ref.@xcite ) .
therefore , in terms of hydrodynamics , we are in the linear regime .
it means that friction force should be linear in the macroion velocity and we expect the average drift speed to be given by the force balance condition , @xmath75 , where @xmath76 is the effective net charge of the macroion complex and @xmath77 is the hydrodynamic friction coefficient .
we shall discuss later the possibilities of determination of the effective net charge @xmath78 based on this condition .
[ fig.t_hist ] also shows significant temporal fluctuations in the drift speed .
inspection reveals that they are larger than what one expects for random kicks of neutral particles .
these fluctuations indicate that neither the counterions permanently stick to fixed points on the macroion nor the coions attach to the counterions , but that they are being replaced from time to time .
the fluctuations of this type are actually seen in a video movie .
the dependence of the average macroion drift speed @xmath79 on the electric field is shown in fig.[fig.vx_e ] . in this figure
, we show together the results of several runs , corresponding to different values of the macroion charge @xmath80 and macroion radius @xmath81 .
first and foremost , the sign of the drift velocity in moderate fields corresponds to the sign of inverted charge .
this is the central observation of our work .
the figure clearly demonstrates the overall pattern of the drift velocity dependence on the applied field , beginning with the linear regime in a weak field , followed eventually by a breakdown of charge inversion in a strong field .
let us discuss the linear drift regime for small electric fields @xmath82 , where @xmath83 increases linearly with the field strength .
this regime corresponds to the usual ohm s law , where the net charge of the complex is insensitive to the strength of the electric field .
a macroion drifts together with its strongly adsorbed counterions as a complex .
that is , the electric field is not strong enough to affect the binding of counterions to the macroion . at this stage , it is necessary to check the issue of hydrodynamic interactions and their screening . for that purpose ,
we show in fig .
[ fig.mu_l ] the effect of the simulation domain size @xmath27 on the macroion drift speed . by a series of the runs with different domain sizes and under fixed number density of neutral particles and ionic strength ,
we have confirmed that at @xmath84 , which is the domain size for the majority of our simulations , the domain - size dependence is essentially leveled off .
this is to be contrasted with polymer chains @xcite , in which case hydrodynamic interactions lead to the finite size effect essentially linear in @xmath85 .
we have also inspected the fluid flow of neutral particles around the macroion .
when rapid fluctuations are averaged out , this flow field does not exhibit any patterns protruding away from the macroion .
this fact implies that the flow of neutral particles induced by the motion of the macroion and other ions is screened at short distances @xcite .
the saturation of the @xmath85 dependence ( fig .
[ fig.mu_l ] ) and the inspection of neutral particle flow patterns both confirm that hydrodynamic interactions are screened in our system , thus making reliable our simulation approach based on the finite domain and the heat bath .
the thermal bath at the distant boundaries does not affect the measured macroion mobility .
the small electric field regime is characterized by the mobility , @xmath86 .
this quantity is plotted in fig .
[ fig.mu_qsurf ] as a function of the macroion bare charge @xmath87 , or , more specifically , on the surface charge density of the macroion @xmath88 .
let us now discuss a practically important question : can we determine the effective net charge of the macroion complex @xmath89 based on the data of mobility measurements , i.e. , based on the data of fig.[fig.mu_qsurf ] - fig.[fig.mu_salt ] ?
physically , as we have already mentioned , the charge @xmath76 is determined by the force balance condition @xmath75 , or @xmath90 .
therefore , determination of the net charge requires knowledge of both the friction coefficient @xmath77 and the mobility @xmath91 .
importantly , the friction coefficient can not be determined by the usual stokes formula @xmath92 , where @xmath93 is the solvent viscosity .
the problem is that the real friction is enhanced by the screening of hydrodynamic interactions @xcite .
the stokes formula is simply understood by that the friction force in general should be proportional to @xmath94 , where @xmath95 is the length scale over which velocity changes , @xmath96 is the relevant surface area . for the stokes problem
, we have @xmath97 . by contrast ,
when the drift is caused by the action of the electric field on the overall neutral system , the distance @xmath95 becomes much smaller . in the debye - hckel theory
, it turns out to be of the order of the debye screening length , @xmath98 . in this case , the friction coefficient becomes @xmath99 i.e. , it is enhanced by the factor @xmath100 compared to the usual stokesian friction .
although we work under the conditions where the debye - hckel theory is not applicable , and we do not know exactly which length should be there instead of @xmath101 in eq.([eq : friction ] ) , this length is clearly smaller than @xmath102 and independent of @xmath102 . therefore , the friction coefficient is proportional to @xmath103 - unlike more familiar stokes case where it scales like @xmath102 . under usual circumstances where the charge @xmath76 is known , the friction coefficient scaling as @xmath104 implies that mobility @xmath105 is determined by the _ surface charge density _ , not by the charge and the surface area separately .
this fact was known to m.smoluchowski already a century ago @xcite . in our simulation , effective charge is not known _ a priori _ , and the logic needs to be reversed .
figure [ fig.mu_qsurf ] indicates that mobility @xmath106 is essentially a constant when the macroion bare surface charge density is not too small ( @xmath107 ) .
given that @xmath108 , we conclude that the effective bare charge @xmath90 is proportional to the surface area of the macroion ; namely , charge inversion is characterized by the overcharging density .
this agrees with the theory @xcite . in order to perform at least very rough quantitative estimate of charge @xmath89 based on the mobility data , we need to know the stokesian friction coefficient @xmath109 ( or equivalently
, we need to know the viscosity of our model solvent ) .
we measure it in a separate molecular dynamics run , by observing an exponential decay of the macroion velocity starting from @xmath110 for the case without an electric field .
we find @xmath111 for a spherical particle of the radius @xmath66 and @xmath112 for the macroion complex with adsorbed counterions and coions .
these two estimates provide lower and upper bounds for the charge @xmath89 .
assuming @xmath113 and @xmath114 ( saturation regime in fig.[fig.mu_qsurf ] ) yields the inverted charge @xmath115 between @xmath116 and @xmath117 .
this is in rough agreement with the @xmath118 measurements .
special attention must be paid to the small bare surface charge density case , for which fig.[fig.mu_qsurf ] indicates the decrease in the reversed mobility . for some cases ,
the reversed mobility even disappears altogether , changing to normal , non - reversed mobility , @xmath119 when the macroion bare surface charge density decreases to about @xmath120 @xmath121 .
it turns out that this is the manifestation of correlations between @xmath48-ions in the bulk solution away from the macroion .
indeed , when the macroion is only weakly charged , the correlations of @xmath48-ions in its vicinity are not much stronger than in the bulk , which suppresses charge inversion .
simple estimate shows that , under the conditions when @xmath91 gets small or negative in fig.[fig.mu_qsurf ] , the `` wigner cell '' radius of the @xmath48-ions on the macroion surface , @xmath122 ( which is about @xmath123 for @xmath124 and @xmath67 ) becomes comparable to the average spacing between the @xmath48-ions in the bulk .
the inspection of the radial charge distribution functions around the macroion for this case agrees with this interpretation .
it shows that the counterions are loosely bound to the macroion , and that the coions form relatively strong bonds with the counterions and drift together with them . to examine the above interpretation further , we perform special runs with reduced concentration of the @xmath0 salt
. the results of these runs are shown in fig.[fig.mu_qsurf ] with filled circles and triangles for the @xmath125 @xmath126-ions with @xmath127 90 and 30 negative coions , respectively .
as anticipated , with fewer @xmath48-ions in the bulk , charge inversion is not interrupted even at small macroion bare charges .
we regard this a convincing proof that charge inversion occurs at small concentration of the @xmath48-ions _ despite _ a larger entropy penalty .
figure [ fig.mu_qsurf ] also shows with crosses the data of the control run performed under the condition of the weak electric field without any heat drainage ( see sec.[sec.2.2 ] ) .
these data are within error bars of the cases with the simulated thermal bath .
the dependence of the macroion mobility @xmath106 on the valence of the counterions @xmath126 in fig.[fig.mu_z ] is physically interesting , and also important for application purposes . for the cases shown with filled and open circles ,
the surface charge density of the macroion is chosen nearly the same @xmath128 so that they reside in the saturation regime of fig.[fig.mu_qsurf ] .
we emphasize that the mobility for these cases is given by the same curve .
the mobility of the macroion is _ negative _ when counterions are monovalent @xmath129 , because there is no charge inversion but only regular debye screening .
thus , the charge inversion phenomenon does not occur in the solution of monovalent salt ( provided that the co- and counter - ions have the equal radius ) . for @xmath130 , charge inversion does take place , as manifested by the _ positive _ mobility .
these observations agree with a previous study for planar charged surfaces @xcite .
a remarkable finding is that the mobility here is maximized for the intermediate valence , @xmath131 , unlike the peak inverted charge that accounts for static charge distribution of mainly counterions @xcite .
it is also noted in fig.[fig.mu_z ] that , when the surface charge density is reduced , both the magnitude of reversed ( positive ) mobility and the range of @xmath126 where it occurs shrink as shown by square symbols in the figure .
the mobility for divalent @xmath48-ions is now negative .
this corresponds to the small surface charge density regime @xmath132 in fig.[fig.mu_qsurf ] .
yet , the mobility is maximized for the intermediate valence , @xmath133 in this case . also shown in fig.[fig.mu_z ] are the results of the control runs performed with reduced number of @xmath48-ions ( discussed above in connection with fig .
[ fig.mu_qsurf ] ) .
they again reproduce the optimal valences for charge inversion .
although somewhat speculative , we can try to apply the result of fig.[fig.mu_z ] to explain the electrophoretic mobility measurements of nucleosome core particles in cation solution @xcite .
what was observed is the increase in the magnitude and range of the cation concentration for occurrence of reversed mobility when spermidine salt ( + 3 ) was replaced with spermine salt ( + 4 ) , while charge reversal was not observed for any concentration of mg@xmath134 .
it is worth stressing that nucleosome particles is a complicated system in which many aspects may be important .
what we would like to say here is that our results may be at least one of the factors relevant to the experiments reported in the literature @xcite . the dependence of the macroion mobility on the salt ionic strength , @xmath135 is shown in fig.[fig.mu_salt ] for the counterion valence @xmath67 and the temperature @xmath136 . here
, the ionic strength is related to the debye length by @xmath137 .
the mobility increases quite rapidly for small ionic strength , and is well fit by @xmath138 as shown by a dashed curve .
it is a legitimate concern to ask whether the data of fig.[fig.mu_salt ] are affected by the correlations of @xmath48-ions in the bulk which we discussed in connection with fig .
[ fig.mu_qsurf ] .
the answer is negative ; these data correspond to the saturation regime of fig .
[ fig.mu_qsurf ] .
the distance between @xmath48-ions in the bulk drops to the value comparable to the wigner - seitz cell only when @xmath139 gets as large as @xmath140 or more .
the increase in the mobility with ionic strength contradicts the intuition based on the debye - hckel theory , and deserves a comment .
as it is explained in detail in @xcite and understood by a number of authors cited there , charge inversion itself grows with ionic strength .
this happens because charge inversion is the result of interplay between the repulsion of counterions from each other and the attraction of each of them to its own correlation hole .
the latter occurs at a much shorter distance than the former , and only the repulsion is strongly affected by screening .
this is why the amount of charge inversion , hence the macroion mobility , grows with increasing ionic strength .
it is also worth mentioning that the quick rise in the reversed mobility at small ionic strength in fig.[fig.mu_salt ] agrees with the colloidal mobility measurement for the case with trivalent salt lacl@xmath141 @xcite .
let us now return to fig.[fig.vx_e ] to discuss the regime that is nonlinear in the applied electric field .
as the figure indicates , the charge - inverted shell around the macroion is destroyed for large electric fields . moreover , the critical field @xmath142 at which this happens is independent of the macroion size , which leads us to an estimate @xmath143 this result is quite interesting .
indeed , @xmath144 is the electric field on the macroion surface produced by the macroion bare charge .
why does the critical field scale with the bare charge of the macroion instead of the net charge of the complex ?
the reason is due to correlations between screening ions .
we noted while discussing fig.[fig.t_hist ] that the counterions on the macroion surface are being replaced from time to time .
consider how one @xmath126-ion can depart from the macroion surface .
since this ion is surrounded by a correlation hole on the surface , its departure requires work against the unscreened bare electric field of the macroion as long as its distance from the surface is smaller than the distance between the adsorbed @xmath126-ions .
therefore , departure from the surface becomes possible when the external field becomes comparable with this unscreened field ; the charge - inverted complex is no longer stable at such a field strength . the critical electric field in realistic situations
is estimated to be very large .
for the parameters @xmath145and @xmath146 , the critical electric field becomes as large as @xmath147 , where we take into account the dielectric constant of water @xmath148 @xcite .
although the critical field is large , it gives small energy to the electric dipole of a water molecule ,
@xmath149 : @xmath150 . this verifies the use of the model solvent of neutral particles in the present molecular dynamics simulations . in practice
, the applied electric field is not expected to disrupt the charge - inverted macroion complex .
in this paper , we performed molecular dynamics simulations with the use of neutral - particle solvent , and measured the drift speed of a macroion to obtain its mobility under electrophoresis in a multivalent @xmath151 salt solution .
a weak electric field pulled the macroion complex in the direction determined by the net inverted charge , instead of disrupting it .
the reversed mobility of the complex , @xmath152 , was shown to be nearly constant for the weak electric fields .
we showed the functional dependences of mobility in figs.3 , 5 and 6 of sec.[sec.3 ] , respectively , against the electric field strength @xmath53 , the surface charge density of the macroion @xmath153 , and the valence @xmath126 of the counterions .
the mobility was a function of the surface charge density , instead of the bare charge and radius of the macroion separately .
the reversed mobility increased rapidly with the salt ionic strength @xmath154 as @xmath138 .
interestingly , the reversed mobility took a maximal value at the intermediate valence of the counterions @xmath155 .
we confirmed the screening of hydrodynamic interactions at a few debye length .
no specific flow patterns of neutral particles , which one would expect for the sphere moving in a viscous fluid , were detected around a macroion . in the large field regime , although academic because of its huge value , electrophoresis was strongly nonlinear , and the field stripped the screening counterions off the macroion .
the mobility of the macroion complex dropped significantly from that of the linear regime , and the sign of the mobility flipped back to non - reversed one above the critical electric field , which was approximately half the macroion _ unscreened _ field . in this study , we explicitly simulated the neutral particles of the solvent to produce a reliable hydrodynamic background .
on the other hand , the limits of computational resources prevented us from inclusion of the @xmath4 salt .
the screening in our system was exclusively accomplished by the @xmath0 salt .
for this reason , we are not ready to make a quantitative comparison of our data with the real experiments . the simulation including the @xmath4 salt is currently under way
. the authors are grateful to professor b.shklovskii and dr.t .
nguyen for discussions , and to dr.j.w.van dam for reading the manuscript .
one of the authors ( m.t . )
thanks professor k.kremer and dr.c.holm for collaboration on the pppm algorithm during his stay at the max - planck institut fr polymerforschung ( mainz , 1999 ) under the support of the max - planck society .
the numerical computation was performed with the vpp800/13 of the institute of space and astronautical science ( japan ) , and partly with the origin 3800 of the university of minnesota supercomputing institute .
d.long , j .- l.viovy , and a.ajdari , _ phys.rev.lett . _ * 76 * , ( 1996 ) 3858 .
l.viovy , _ rev.mod.phys . _ * 72 * , ( 2000 ) 813 .
landau and e.m .
lifshitz , _ theoretical physics , vol . 6 , fluid mechanics _ ( butterworth - heinemann , 1990 ) .
e.m.lifshitz and l.p.pitaevskii , _ theoretical physics , vol .
10 , physical kinetics _
( butterworth - heinemann , 1981 ) , chapter 2 .
m.smoluchowski , _ bull .
international de lacademie des sciences de cracovie _ * 8 * , ( 1903 ) 182 - 200 . | we have performed molecular dynamics simulations to study the effect of an external electric field on a macroion in the solution of multivalent @xmath0 salt . to obtain plausible hydrodynamics of the medium
, we explicitly make the simulation of many neutral particles along with ions . in a weak electric field
, the macroion drifts together with the strongly adsorbed multivalent counterions along the electric field , in the direction proving inversion of the charge sign .
the reversed mobility of the macroion is insensitive to the external field , and increases with salt ionic strength .
the reversed mobility takes a maximal value at intermediate counterion valence .
the motion of the macroion complex does not induce any flow of the neutral solvent away from the macroion , which reveals screening of hydrodynamic interactions at short distances in electrolyte solutions . a very large electric field , comparable to the macroion unscreened field , disrupts charge inversion by stripping the adsorbed counterions off the macroion . |
semi - classically black holes are known to exhibit no - hair @xcite and exact thermality of hawking evaporation @xcite . despite the fact that these properties are derived in an idealized semi - classical limit in which black holes are infinitely heavy and infinitely long - leaved
, one usually assumes that these are good ( up to exponentially small deviations ) approximations for real black holes of finite mass and lifetime , as long as they are much heavier than the planck mass ( @xmath2 ) .
such extrapolation then leads to highly profound restrictions on the symmetry and the particle content of the theory .
the two examples of such restrictions are : 1 ) the folk theorems " that state incompatibility of exact global symmetries with gravity ; and 2 ) the restriction on the number of light particle species , @xmath0 @xcite . both restrictions , when quantified lead to surprisingly similar constraints in the following sense . let us denote by @xmath3 the maximal possible dimensionality of a representation under a global symmetry group allowed in a given theory .
essentially @xmath3 counts the maximal amount of the measurable global charge in a given theory . for continuous symmetries
@xmath3 is infinite , since a global charge ( e.g. , @xmath4 baryon number ) can be increased without a limit by considering the states with arbitrary large occupation number of the charged particles ( e.g. , baryons ) .
let us denote by @xmath0 the number of particle species in the theory
. for definiteness let us assume the species to be massless .
of course @xmath3 and @xmath0 are unrelated quantities .
one counts the number of species , whereas the other counts a maximal charge , which can be arbitrarily large even if @xmath5 . despite this difference , in both cases , under the assumption of the validity of semi - classical properties to finite - mass black holes , one can design thought experiments that lead to the emergence of very similar length - scales below which the theory must be modified . for the species case
the scale is @xcite @xmath6 whereas for global ( or discrete gauge ) symmetries the scale reads @xcite @xmath7 where @xmath8 is the planck length that in terms of the planck mass and newton s constant can be expressed as @xmath9 . in the above semi - classical treatment ,
both scales appear to have universal and fundamental meaning , since these are the shortest scales beyond which black hole no - hair and thermality properties must be fully violated .
thus , both scales are treated as universal cutoff for applicability of einstein gravity .
both constraints would have far - reaching consequences .
for example , ( [ lglobal ] ) immediately implies impossibility of having exact continuous global symmetries , since in such a case one would have @xmath10 which would imply @xmath11 .
however , it is crucial to remember that in attributing a fundamental meaning to the scales @xmath12 and @xmath13 , one is based on the very strong assumption that semi - classical properties can be extended to real ( quantum ) black holes of finite mass . whether this is a valid assumption
can only be answered in a microscopic fully - quantum description .
this question was addressed in @xcite within the context of the recently - formulated quantum portrait of black holes @xcite .
it was shown there that the extension of semi - classical no - hair arguments to real black holes was incorrect , since black holes do carry hair even under global charges to the order @xmath14 , where @xmath15 is the occupation number of gravitons within the black hole , and @xmath3 is the amount of the global charge swallowed by the black hole . as a result ,
the basic assumptions leading to `` folks theorems '' about the non - existence of global charges are false . within the black hole quantum portrait ,
the scale @xmath12 is fully reproducible , however it has no universal meaning .
it simply marks a stage in black hole s evolution history beyond which a particular black hole that initially swallowed @xmath3 units of a global charge depletes to the point where occupation numbers of gravitons and global charges equilibrates , i.e. @xmath16 . hence at this state , the global hair becomes fully important . at the same time
a different black hole that was endowed by a zero global charge , will continue an almost - thermal depletion all the way till it reaches the size @xmath17 .
so far the discussion was constrained to the pure gravity case .
however particle species , which are coupled to gravitons , exist in nature .
therefore in the present paper we shall extend the analysis of @xcite applying it to the particle species hair .
in particular , we will investigate the quantum meaning of the scale @xmath13 .
we shall discover that unlike @xmath12 , the scale @xmath13 has an universal and a fundamental meaning .
it is a scale beyond which an _ arbitrary _ black hole in a given theory , regardless of its initial composition , ceases to deplete at an approximately - thermal rate . moreover , in full accordance with @xcite , @xmath13 is a lower cutoff beyond which the species identities can not be resolved .
as we shall see , the black hole quantum portrait reveals an important information about the underlying quantum physics behind this scale , which is impossible to read - off in the standard semi - classical treatment .
namely , the meaning of species scale is to set the size of lightest quantum black holes existing in the theory .
our findings have a number of important implications .
first , in full accordance with @xcite it nullifies the so - called species problem ( for a review of species problem , see @xcite ) .
one incarnation of this problem is a seeming inconsistency between the bekenstein and entanglement @xcite entropies in the presence of many particle species .
our result confirms the claim of @xcite , that this problem is an artifact of not taking into the account the correct cutoff of the theory which is given by @xmath18 rather than @xmath2 .
secondly , our findings have direct implication for the concept of _ self - uv - completion _ of gravity by black hole physics @xcite .
as explained in @xcite for self - completion of gravity it is essential that the first quantum black holes exist right above the scale of perturbative unitarity violation . in this way , the black hole physics takes over and theory can classicalize exactly above the scale where perturbative unitarity fails .
as argued in @xcite , in pure einstein gravity the two scales meet at @xmath2 , and the existence of the lightest quantum black holes follows from the existence and the depletion of the heavier ones . in the language of the quantum @xmath15-portrait this claim becomes explicit in terms of occupation number of gravitons .
the lightest quantum black holes are the ones with @xmath19 .
their mass is @xmath2 , which coincides with perturbative scale of unitarity violation .
how the presence of species changes this balance ?
as already suggested in @xcite the result must be sensitive to the nature of species . if species are gravitational ( such as e.g. , kaluza - klein gravitons ) the property of having lightest black holes exactly at the unitarity violation scale does not change .
on the other hand non - gravitational species create a gap between the two scales .
our present analysis within the quantum portrait exactly confirms this picture .
we see , that gravitational species , equally affect the depletion as well as self - sustainability properties of the graviton condensate . as a result ,
the lightest quantum black holes ( @xmath20 ) have the mass @xmath18 , which as we shall show exactly coincides with the scale of perturbative unitarity violation .
thus , in this case , the black hole classicalization regime smoothly merges with the perturbative unitarity regime without any gap .
in contrast , when species are non - gravitational the balance is violated .
non - gravitational species ( similarly to gravitational ones ) change the depletion law of the condensate , but fail to change the self - sustainability properties . as a result , we get a finite gap of order @xmath21 between the scale of perturbative unitarity violation and the first quantum black hole . generalizing this result for arbitrary mix of gravitational and non - gravitational species , we reach the following powerful conclusion .
@xmath22 _ the absence of a gap between the scale of perturbative unitarity violation and the lightest quantum black holes demands that the number of non - gravitational species should not exceed the number on gravitational ones .
_ @xmath22 taking the idea of self - completion seriously , the existence of about @xmath23 non - gravitational species in the standard model can be taken as indication of at least equal number of gravitational ones below the scale of @xmath24gev .
the paper is organized as follows . in the next section
we will generalize the _
black holes s quantum n - portrait _ including besides the gravitons also other particle species into the black hole bound state . as a result
we will see that the species scale is a fundamental , universal length in the quantum picture .
we then give a physical meaning to this scale in terms of measurement of species identities .
finally , we discuss the crucial difference between gravitational and non - gravitational species , and the connection with self - uv - completion of gravity . throughout the paper in the equations
we shall ignore the order one numerical factors , which are unimportant for our analysis .
before proceeding , let us set the stage by briefly formulating main aspects of the black hole quantum theory @xcite . for a detailed discussion the reader is referred to the original papers . according to this picture ,
in pure gravity , i.e. in the absence of other species , the black holes represent a self - sustained bound - states ( a bose condensate ) of weakly interacting long - wavelength gravitons .
the remarkable thing about this bound - state is that the condition of self - sustainability can be satisfied for an arbitrary occupation number @xmath15 and fixes the characteristic wavelength of gravitons as @xmath25 .
all the other characteristics of the system , such as the total mass ( @xmath26 ) and the coupling of individual gravitons ( @xmath27 ) are uniquely determined by @xmath15 as @xmath28 and total mass @xmath29 .
relations in t hooft s sense @xcite . ] in this picture , we operate exclusively by quantum mechanical notions , such as quantum coupling , wave - length and occupation number . all the geometric and thermodynamical characteristics , such as horizon and temperature ,
are emergent as a result of semi - classical limit which corresponds to double scaling limit @xmath30 the notions of the geometric schwarzschild radius @xmath31 and of the temperature @xmath32 only emerge in this limit .
thus , the quantity @xmath31 , which quantum - mechanically is a characteristic wave - length of the condensate , acquires a geometric meaning only in the semi - classical limit .
not surprisingly many of the intrinsically - quantum properties ( such as black hair ) become lost in this limit , unless one carefully monitors their scaling also .
the above mentioned `` folk theorems '' about violation of global charges are precisely an artifact of not taking into the account the existence and scaling of the black hole hair @xcite .
let us now assume that besides the graviton , the theory includes other massless species with total number of species being @xmath0 .
let us see how the existence of the extra species in the theory spectrum affects the depletion law .
we start with a self - sustained condensate that is initially composed only out of @xmath15 gravitons . in
the absence of other species , the leading contribution to the depletion rate is coming from the process in which two graviton re - scatter in such a way that one of them gains the above - escape energy and leaks out of the condensate .
the corresponding feynman diagram is depicted in fig.1 0.5 cm the rate of this process goes as , @xmath33 which gives the following depletion equation , @xmath34 notice that the quantity , which in the semiclassical limit acquires the meaning of hawking temperature , is @xmath35 in terms of this parameter the depletion equation can be written as @xmath36 rewriting @xmath15 in terms of the black hole mass the depletion equation becomes a stefan - boltzmann law for a black hole , @xmath37 thus , the depletion imitates an approximately - thermal evaporation all the way until @xmath15 becomes order one , or equivalently , until the characteristic wave - length of the graviton condensate becomes of order @xmath38 .
let us now see how the presence of extra species changes this situation . in the presence of additional species ,
the contribution to the rate is enhanced by the processes , in which two gravitons now can annihilate also into the pair of two species .
the corresponding diagram is shown in fig .
2 . as a result
the depletion rate is now enhanced by factor of @xmath0 , @xmath39 which gives the following depletion equation , @xmath40 notice , that the quantum precursor of temperature is still given by the quantity ( [ t ] ) . taking the time - derivative of this equation and using ( [ depletens ] ) , we arrive at the following relation , @xmath41
this equation tells us that as soon as the number of gravitons in the condensate depletes to become comparable to @xmath0 , the quantity @xmath42 can not even approximately be interpreted as a temperature , since the rate of its change exceeds its own value .
the corresponding wavelength of the graviton condensate is given by , @xmath6 thus , we uncovered the quantum meaning of the species scale as the wave - length of the self - sustained condensate , for which @xmath43 .
this scale is an _
universal _ characteristics of a given theory .
any black hole , irrespective of its initial composition , will loose even an approximate thermality ones its size will reach @xmath13 .
we thus see that the scale @xmath13 maintains its universal meaning even in the quantum picture .
it is also evident that this scale must have some fundamental meaning in the sense that the black holes must undergo a qualitative change in their nature beyond this scale .
this is obvious from the fact that it makes no sense to talk about depleting self - sustained bound - state of gravitons beyond this point , because the depletion time is much shorter than the wave - length .
moreover , let us show that we can never reach purely gravitational bound - states of sizes smaller than @xmath13 , even if we start out with a large black hole composed exclusively out of gravitons . in order to see this
let us follow the evolution of a black hole which is initially composed only out of gravitons .
the leading process of the depletion converts the black hole made out of @xmath15 gravitons into the one with @xmath44 gravitons and one new species .
so the black hole gets endowed with species hair .
the probability of getting rid of this hair is suppressed by a factor @xmath45 relative to the probability of decreasing the number of gravitons .
so after every step the occupation number of species in the black hole increases with the chance @xmath0 to one . this process shall continue until the number of other species and gravitons in the black hole equilibrates .
moreover , the numbers of species and their anti - particles must be roughly the same .
after this point the evolution of the condensate depends on the details of interaction between the species .
but , in any case the departure from thermality for such a condensate is maximal .
the study of its dynamics is a very interesting problem per se , but beyond the scope of the present paper .
the lesson we would like to draw from this analysis is that it is highly unprobable that a purely graviton self - sustained bound - state can survive going through the scale @xmath13 .
thus the quantum portrait gives a simple microscopic rational for the universal nature of the species scale @xmath13 .
finally , let us discuss another fundamental meaning of the species scale .
as suggested in @xcite within the semi - classical treatment , the species scale represents a lower bound on the scale down to which the species identities can be resolved .
let us see how this argument goes through the quantum portrait .
the semi - classical argument of @xcite is structured as follows .
imagine that we would like to build a smallest pixel that could resolve the species identities in shortest possible time ( that is , on the time - scale of its size ) .
let the size of the pixel be @xmath46 .
the necessary condition that such a pixel must satisfy is to have localized sample of at least one member from each species ( or an equivalent information ) .
thus , we have at least @xmath0 particles localized within the pixel .
the process of resolving the species identities is then a scattering process of an unknown particle at the pixel ( see fig .
3 ) . this requirement gives the pixel a minimal mass , @xmath47 then any attempt of decreasing the size of the pixel below @xmath13 would fail since it would result into the formation of a black hole of schwarzschild radius exceeding @xmath13 .
we thus arrive to the lower cutoff being @xmath13 .
let us now reinterpret this reasoning in our quantum language .
quantum mechanically the pixel is a bound - state consisting of at least @xmath0 particles . in reality
, one needs an extra interaction to localize the species wave - functions within @xmath46 , which will increase the energy of the pixel beyond ( [ pixm ] ) ; however this only works in favor of our bound , since it increases the chances of black hole formation .
now , according to the quantum portrait @xcite , the energy of localized species necessarily results into the occupation number of gravitons , given by @xmath48 , which is convenient to rewrite as @xmath49 the latter form is very instructive , since it indicates that as soon as we try to decrease @xmath46 below @xmath13 the occupation number of gravitons will exceed the number of species , and moreover the gravitons will form a self - sustained bound - state , a black hole with the occupation number of other species being a small fraction of the number of gravitons , @xmath50
. the wavelength of the resulting graviton bound - state ( a size of a formed black hole ) will be @xmath51 of course , this black hole carries a detectable species hair , but the probability of extracting it is @xmath52 suppressed per species and thus takes much longer than the time @xmath13 .
namely , to detect a hair of each individual species takes the time @xcite @xmath53 thus , we see why the scale @xmath13 is indeed a lower bound on the resolution capacity of the species identities .
the above reasoning shows that the scale @xmath13 is fundamental even in quantum - mechanical sense . in particular
, all the virtual processes in which species counts differently must be cutoff at this scale .
thus , in a theory with species the cutoff scale is @xmath13 and not @xmath38 .
necessity of such a cutoff immediately explains why there is no problem of in counting of the entanglement entropy in semi - classical limit @xcite .
so far we were not interested in the nature of species
. we shall now become more specific and try to distinguish between the _ gravitational _ and _ non - gravitational _ species .
as we shall see , the behavior of smallest black holes is dramatically different in the two cases . in case of gravitational species
the mass of the lightest black holes ( `` planckions '' ) is right at the scale at which the perturbative violation of unitarity takes place .
as it was stressed in @xcite , this property is crucial for the self - completeness of theory , since the unitarity is restored by quantum black holes .
in contrast , in case of non - gravitational species there is a mass gap between the scale of perturbative unitarity - violation and the first available quantum black holes .
uv - completion of theory demands that this gap be filled up by some non - black hole physics .
let us now highlight this fundamental difference in more details .
we shell define _ gravitational species _ as the ones that just like graviton are sourced by the energy momentum of the existing particles in the theory .
essentially , this set includes all the elementary spin-@xmath54 species that may exist in the theory . of course , by general covariance all such states except one ( the `` true '' graviton )
must be massive , and their number will obey a certain distribution .
we shall parameterize this distribution by denoting @xmath55 the number of all species at and below a mass level @xmath56 .
a familiar example of such a distribution is provided by kaluza - klein theories , although we shall keep our discussion independent of geometric considerations .
the crucial property of gravitational species is that they not only affect the depletion law , but also the self - sustainability condition of the condensate . in order to make this clear
, we shall work in a particular example of _ extreme democratic _ sourcing , in which all the species universally source each other with the same strength as the massless graviton , @xmath57 , where @xmath58 is a characteristic wave - length in the process .
of course , in concrete cases the strength and the type of graviton couplings can ( and will ) obey certain selection rules , as it happens for example in kaluza - klein theories , however for our purposes this idealized approximation is good enough as it maximally sharpens the point . in such a situation for the gravitons in a self - sustained bound - state of wave - length @xmath58
the effective gravitational coupling becomes @xmath59 this fixes the self - sustainability relation between the wave - length and the occupation number @xmath15 as , @xmath60 where the @xmath61 has to be understood as `` running '' species scale due to the wave - length dependence of the number of species involved .
the depletion rate in such a case is @xmath62 where the @xmath63 factor comes from the fact that , due to democratic inter - sourcing , each constituent pair can annihilate into @xmath63 pairs , as it is depicted on fig .
4 . taking into the account ( [ wave ] ) , the above rate can be rewritten is a very simple form , @xmath64 comparing with ( [ leakage ] ) we see that depletion law is the same as for a single species , but with @xmath38 being replaced by @xmath13 .
this is an extremely important fact , which tells us that for gravitational species the depletion law remains approximately thermal all the way till @xmath20 .
notice , that the state @xmath20 , is a lowest mass member of the black hole tower .
the mass and the characteristic wavelength of such a lightest black hole is , @xmath65 notice that this coincides with the scale of perturbative unitarity violation in the theory .
indeed , consider a tree - level @xmath66 scattering process depicted in fig .
5 , in which a two initial particles annihilate into any of the two species .
scattering process of gravitational species responsible for violation of perturbative unitarity above @xmath18 .
, width=434 ] at center of mass energy @xmath67 , the rate of the process goes as @xmath68 where the @xmath63 factor appears due to summation over both internal ( green ) and external ( red ) lines .
obviously , this process violates perturbative unitarity at the energy @xmath69 comparing with ( [ mini ] ) we see that the quantum black holes enter into the game right above the unitarity violating scale @xmath13 .
the picture is essentially the same as for single species , with the only difference that everything happens at a lower energy scale , with @xmath13 which takes up the role of @xmath38 .
thus , the necessary condition for self - completion @xcite is satisfied .
the quantum black holes are readily available in the spectrum right above the unitarity scale !
it is obvious that situation is very different for non - gravitational species .
such species affect the depletion law , but at the same time do not make binding force stronger . as a result , the black hole goes out of thermal regime while still being a multi - particle state , much heavier than the perturbative - unitarity violation scale .
indeed , since non - gravitational species only affect the depletion law and not the self - sustainability condition of the graviton condensate .
the mass of the smallest black hole in such a case is @xmath70 .
notice that this exceeds the perturbative unitarity violating scale , which for non - gravitational species is at least as low as , @xmath71 this unitarity bound comes from a @xmath66 scattering process between the species via graviton exchange , depicted in fig.6 .
the rate of this process goes similar to eq.(22 ) , except there is one less power of @xmath0 due to the fact that there is only one intermediate graviton exchange .
so the question is what fills the unitarity gap between this scale and the smallest black holes ?
one way or another , we need some non - black hole physics for curing the unitarity problem in the interval of energies @xmath72 this situation is schematically depicted in fig 7 .
the red lines represented regions of perturbative unitarity , whereas the green ones the regions in which unitarity is restored by quantum black holes .
the dashed region in the left is a non - unitarity window in case of non - gravitational species .
@xmath22 our analysis can be easily generalized to a mixed situation when species consist of both gravitational and non - gravitational ones .
the particular details of course change , but the following conclusion persists . whenever number of non - gravitational species is much larger than the gravitational ones , there is gap between the unitarity scale and the masses of the smallest black holes .
this suggest the following surprisingly powerful conclusion : _ self - uv - completion of gravity demands that the number of non - gravitational species be less or equal than the gravitational ones ( normalized per strength of einstein s
graviton)_. in this paper we have reconsidered the semi - classical black hole bound on the number of particle species @xcite within the microscopic quantum portrait and have shown that the bound persists in full quantum picture .
introduction of @xmath0 in the theory creates a new fundamental length scale , @xmath13 , which is an ultimate lowest cutoff on the black hole size . within quantum portrait
this scale acquires a very clear physical meaning , as the characteristic black hole size ( or to be more precise , the characteristic wave - length of the black hole constituents ) beyond which the depletion process _ not even approximately _ can mimic the hawking thermal spectrum .
essentially , no black holes of the smaller size exist in a given theory .
furthermore , we have shown that physics of the cutoff is highly sensitive to the nature of species .
for gravitational species , the scale @xmath18 simultaneously marks the upper bound on the scale of perturbative unitarity , as well as the lowest bound on the black hole mass .
this is a strong indication in favor of self - uv - completion of gravity by classicalization through quantum black holes .
indeed , the quantum black holes that are crucial for unitarizing scattering amplitudes are readily available right at the scale of perturbative unitarity violation .
situation is dramatically different for non - gravitational species ( such as e.g. , extra neutrino species ) . in this case
there is a mass gap between the mass of the lightest black holes and the scale of perturbative unitarity . this gap ,
unless filled by some yet unknown physics , indicates an intrinsic inconsistency of the theory .
we are thus , lead to the conclusion that in any consistent theory of gravity the number of gravitational species must be the dominant one .
all the known examples of deformation of einstein gravity , such as kaluza - klein or perturbative string theory , satisfy this property .
the work of g.d . was supported in part by humboldt foundation under alexander von humboldt professorship , by european commission under the erc advanced grant 226371 , by trr 33 the dark universe and by the nsf grant phy-0758032 .
the work of c.g . was supported in part by humboldt foundation and by grants : fpa 2009 - 07908 , cpan ( csd2007 - 00042 ) and hephacos p - esp00346 .
this work of d.l . is supported by the dfg transregional collaborative research centre trr 33 and the dfg cluster of excellence ` origin and structure of the universe ' .
g. dvali , `` black holes and large n species solution to the hierarchy problem , '' arxiv:0706.2050 [ hep - th ] ; g. dvali and m. redi , `` black hole bound on the number of species and quantum gravity at lhc , '' phys .
* d77 * ( 2008 ) 045027 , arxiv:0710.4344 [ hep - th ] .
g. dvali and c. gomez , `` black hole s 1/n hair , '' arxiv:1203.6575 [ hep - th ] . g. dvali , c. gomez , black hole s
quantum n - portrait , arxiv:1112.3359 [ hep - th ] ; landau - ginzburg limit of black hole s quantum portrait : self similarity and critical exponent , arxiv:1203.3372 [ hep - th ] .
w. israel , phys .
lett . a57 ( 1976 ) 107 ; l. bombelli , r. k. koul , j. h. lee and r. d. sorkin , phys .
d 34 , 373 ( 1986 ) ; m. srednicki , phys . rev .
71 , 666 ( 1993 ) ; v. p. frolov and i. novikov , phys . rev .
d 48 , 4545 ( 1993 ) . | recently within the context of a microscopic quantum theory , the _ black hole s quantum n - portrait _ , it was shown that continuous global symmetries are compatible with quantum black hole physics . in the present paper
we revise within the same framework the semi - classical black hole bound on the number of particle species @xmath0 .
we show that unlike the bound on global charge , the bound on species survives in the quantum picture and gives rise to a new fundamental length - scale , @xmath1 , beyond which the resolution of species identities is impossible .
this finding nullifies the so - called species problem .
this scale sets the size of the lightest quantum black hole in the theory , _
planckion_. a crucial difference between the gravitational and non - gravitational species emerges . for gravitational species ,
the lightest black holes are exactly at the scale of perturbative unitarity violation , which is a strong indication for self - uv - completion of gravity .
however , non - gravitational species create a gap between the perturbative unitarity scale and the lightest black holes , which must be filled by some unitarity - restoring physics .
thus , self - uv - completion of gravity implies that the number of non - gravitational species must not exceed the gravitational ones . |
wave function collapse models ( cms ) @xcite , as the modified theories of standard quantum mechanics , postulate a fundamental breakdown of quantum superposition at the macroscopic scale .
they have been proposed to explain the lack of observations of macroscopically distinguishable superposition states of macroscopic objects and offer a possible solution to the quantum measurement problem @xcite and the quantum - to - classical transition .
there are several different versions of the cms , e.g. , the ghirardi - rimini - weber approach @xcite , continuous spontaneous localization ( csl ) @xcite , and gravitationally - induced cms @xcite .
these models modify the schrdinger equation by introducing appropriate stochastic nonlinear terms , of which the effect is negligible for microscopic systems , while becomes prominent for macroscopic objects resulting in the emergence of macroscopic classicality .
many proposals have been put forward to examine these cms in different systems and with different approaches . in general , they can be divided into two kinds : in an interferometric way and in a noninterferometric way . for the former ,
matter - wave interferometry @xcite is typically used , where large massive objects , e.g. , molecules or clusters @xcite , are sent through interference gratings .
however , to date the mass range has not been reached to effectively test the cms . for the latter , protocols @xcite have been recently provided based on cavity optomechanics @xcite .
the main advantage of this approach is that the preparation of large spatial superposition states @xcite is not required .
it has been shown @xcite that collapse noise induced momentum diffusion of a mechanical resonator ( mr ) could be probed in the phase noise of the cavity output light . apart from the above , more recently quantum estimation theory has been applied to discriminate the effect of cms @xcite . in this paper
, we provide a novel scheme to test the cms by means of the steady - state entanglement of two macroscopic mrs .
many protocols have been proposed for the generation of entanglement between two massive mrs using optomechanical and/or electromechanical systems , e.g. , by exploiting radiation pressure @xcite , by transferring entanglement @xcite or squeezing @xcite from optical fields , by conditional measurements on light modes @xcite , and by reservoir engineering realized by properly choosing multi - frequency drivings @xcite .
we will focus on the csl model which is one of the most widely studied cms .
the scheme is based on the known fact that entanglement , as a kind of quantum correlations , is particularly sensitive to various noises .
a small rise of noise may significantly degrade the entanglement .
as is known , the collapse noise postulated in the cms is typically very small thus bringing the challenges for experimental verification . in view of these , entanglement could act as a perfect probe that may be able to sense whether the collapse noise is present or not .
the reason we adopt _ steady - state _ entanglement is that , in order to efficiently test the cms , one should prepare entangled states lasting for a relatively long time @xcite .
we know that entanglement in steady states , rather than in transient states , is harder to prepare due to the continuous decoherence process interacting with various noises .
the effect of the cms would be more noticeable in the steady - state entanglement due to the time accumulation effect .
the system used to test the collapse theories should possess as small as possible environmental noises , comparable to the hypothetical collapse noise , and the test object should be large or massive enough yielding considerable collapse effect .
levitated nanospheres @xcite , owning very high mechanical quality factors , could be the ideal platform to implement such a test .
protocols @xcite have been provided using a single nanosphere trapped in a fabry - prot optical cavity that are able to test the strength of the collapse rate in the csl theory to values as low as @xmath0 hz with realistic parameters . in the present paper , here we study the effect of the csl on quantum correlations of two macroscopic objects . specifically
, we employ two nanospheres optically trapped in a fabry - prot cavity and study the csl effect on the stationary entanglement of the center - of - mass motion of the two spheres .
since the entanglement is particularly sensitive to noises , the diffusion rates of various noises in the system must be small .
we find that , with properly chosen parameters , the collapse noise results in large reduction of the entanglement and the entanglement shows distinguishable scalings , with and without the csl effect , with certain system parameters , e.g. , the trapping frequency , due to the fact that different sources of noise exhibit different scalings with the system parameters @xcite .
the above observation can unambiguously determine whether the collapse noise is or is not actually present or effective .
the remainder of the paper is organized as follows . in sec .
[ system ] we describe in detail our system , provide the quantum langevin equations for achieving the entanglement of the two nanospheres , and analyze the diffusion rates of various noises ( both environmental and the postulated collapse noise ) in the system . in sec .
[ results ] we present the results and show the details of the parameters with which the csl effect with the corresponding value of the collapse rate could be determined . finally , sec .
[ concl ] is devoted to conclusions and discussions and the appendix gives the details for calculating the steady - state entanglement . and
@xmath1 in a fabry - prot cavity .
the cavity mode is bichromatically driven at the two frequencies @xmath2 and @xmath3 .
large and robust entanglement between the two spheres can be generated in the steady state .
the output field of the cavity is fed back into the input port through highly reflective mirrors ( hrm ) and a controllable beamsplitter ( cbs ) with tunable reflection coefficient @xmath4 , which is used for reducing the cavity loss and improving the entanglement.,width=316 ]
before making the test , we shall first prepare two nanospheres into entangled states .
we adopt the scheme provided in ref .
@xcite which is able to generate large and robust entanglement between two mrs in steady states . as depicted in fig .
[ fig1 ] , we consider two identical spheres of radius @xmath5 trapped by harmonic dipole traps with different frequencies @xmath6 and @xmath1 in two different potential wells within a fabry - prot cavity . the center - of - mass motion of the sphere is modeled as a quantum - mechanical harmonic oscillator .
the cavity is of length @xmath7 , finesse @xmath8 , and mirror radius of curvature @xmath9 . a single cavity mode with resonance frequency @xmath10 , interacting via the usual optomechanical interaction with the two mrs ,
is bichromatically driven , with powers @xmath11 and @xmath12 , at the two sideband frequencies @xmath13 and @xmath14 with the reference frequency @xmath15 detuned from the cavity resonance by @xmath16 .
this means that the cavity mode is simultaneously driven at the blue sideband associated with the 1st mr and at the red sideband associated with the 2nd mr .
we note that the scheme @xcite is the improved version of the one @xcite by introducing a coherent feedback loop , which leads to a significantly reduced effective cavity decay rate and a remarkable improvement of the entanglement .
this is vital and makes it possible to test the cms based on this scheme because the mechanical frequency @xmath17 now can take much smaller values ( since the effective cavity decay rate @xmath18 is significantly reduced due to the feedback ) to fulfill the condition of the scheme @xmath19 @xcite . as shown in ref .
@xcite , the diffusion rate @xmath20 due to the scattering of trapping light , which will be the main diffusion for the system at low pressure and temperature , is proportional to the trapping frequency @xmath17 . a much smaller @xmath17 yields a much smaller diffusion rate @xmath20 , making it possible to generate sizable stationary entanglement between the two mrs
. 1.1cm*(a)*5.28cm*(b)*5.25cm*(c ) * + 0.1 cm 0.1 cm + the system dynamics can be efficiently studied by linearizing the optomechanical interaction in the limit of large driving field .
the relevant degrees of freedom for the linearized dynamics are the fluctuations of the cavity field and of the mechanical center - of - mass variables about their respective average values , described by the amplitude and phase quadratures @xmath21 and @xmath22 ( with @xmath23{=}{{\rm i}}$ ] ) of the cavity field , and by the dimensionless position and momentum @xmath24 and @xmath25 ( with @xmath26{=}{{\rm i}}$ ] , @xmath27 ) of the nanospheres .
the corresponding quantum langevin equations , in the interaction picture with respect to @xmath28 , are given by @xcite @xmath29 where we have set the effective detuning @xmath30 ( @xmath31 the frequency shift due to the optomechanical interaction and also the feedback ) equal to zero , which is the optimal detuning for the generation of entanglement @xcite .
@xmath32 is the effective cavity decay rate due to the coherent feedback where @xmath4 is the reflection coefficient of the controllable beamsplitter in the feedback loop ( see fig . [ fig1 ] ) , @xmath33 is the phase shift of the light in the feedback loop , and @xmath34 ( @xmath35 the speed of light ) is the cavity decay rate without feedback .
we see that @xmath18 can be significantly reduced when @xmath36 and @xmath37 ( @xmath38 ) . in practice
, one can always adjust this phase shift and set it equal to the optimal value for the entanglement @xmath37
@xmath39 is the effective optomechanical coupling where @xmath40}$ ] and @xmath41 @xcite is the bare optomechanical coupling associated with the @xmath42-th mr , with @xmath43 the cavity wavelength , @xmath44 the electric permittivity of the sphere , @xmath45 its volume , and @xmath46 the cavity mode volume with mode waist @xmath47 .
@xmath48 is the mechanical damping rate due to the friction with residual air molecules , with @xmath49 the gas pressure , @xmath50 the mass density of the sphere , @xmath51 the mean speed of the air molecules , @xmath52 their mass ( which we take @xmath53 amu ) , and @xmath54 the gas temperature @xcite . for levitated nanospheres @xmath55 can be very small leading to very high quality factors , @xmath56 @xcite . @xmath57 and @xmath58 are the quadratures of the vacuum noise entering into the cavity and their only nonzero correlation functions are @xmath59 @xmath60 and @xmath61 are the combined force operators in the rotating frame which include all the relevant stochastic forces accounting for the mechanical diffusion .
the only nonzero correlation functions are @xmath62 where @xmath63 , @xmath64 , @xmath65 and @xmath66 are , respectively , the diffusion rates caused by the scattering of background air molecules , of trapping light , of cavity photons , and by the collapse noise . in the relevant high temperature limit , @xmath63 is given by @xmath67 the diffusion rates due to the scattering of trapping and cavity light are given , respectively , by @xcite @xmath68 where @xmath69 , @xmath70 , @xmath71 is the frequency of the trapping laser , @xmath72 is the intensity of the trapping field , which is given by @xmath73 with @xmath74 the laser power and @xmath75 the waist of trapping light which is approximated by @xmath76 with @xmath77 the numerical aperture , and the trapping frequency is determined by @xmath78^{1/2}$ ] . finally , @xmath79 is the mean cavity photon number . according to the csl theory ,
the collapse noise induced diffusion rate for a spherical particle is given by @xcite @xmath80\frac{r_c^4}{r^3 } \\ , \label{lambda_sph}\ ] ] with @xmath81 the atomic mass unit . the actual strength of collapse noise
is determined by two phenomenological parameters : the characteristic length @xmath82 and the collapse rate @xmath83 .
the characteristic length is typically set at @xmath84 nm , above which the collapse effect tends to be prominent .
@xmath83 denotes the average collapse rate at one proton mass .
the initial estimate of @xmath83 is @xmath85 hz @xcite , while larger values have been proposed , e.g. , @xmath86 hz given by adler @xcite . up to now , different experiments have indicated that @xmath83 should be lower than @xmath87 hz @xcite , @xmath88 hz @xcite and @xmath89 hz @xcite , for @xmath84 nm . does it exist an exact value or range of @xmath83 and , if exists , how large is it ?
these questions can only be answered by experiments . 0.8cm*(a)*5.2cm*(b)*5.2cm*(c ) * + 0.1 cm 0.34 cm 0.315 cm + 0.8cm*(d)*5.2cm*(e)*5.2cm*(f ) * + 0.1 cm 0.1 cm +
.different scalings of the diffusion rates @xmath20 , @xmath90 , @xmath91 and @xmath92 with two key parameters of the system , i.e. , the radius of the sphere @xmath5 and the trapping frequency @xmath93 . symbol @xmath94 ( @xmath95 ) denotes an increasing ( constant ) function of the parameter . [ cols="<,<,<,<,<",options="header " , ] we observe that @xmath20 , @xmath96 ( @xmath90 is , however , independent with @xmath5 for fixed values of @xmath97 ) , @xmath92 increases with @xmath5 when @xmath98 , while @xmath99 , as displayed in table [ table ] . since the entanglement is particularly sensitive to various noises , implying that a relatively small size of the sphere should be adopted , we consider the radius of the sphere @xmath100@xmath101 @xmath82 ( while a larger size @xmath102 has been used in ref .
@xcite where the observable is instead the phase quadrature of the output light ) , which yields not large @xmath20 , @xmath90 , @xmath91 and meantime comparable @xmath92 , as shown in fig .
[ fig2 ] ( a ) .
we have taken as small as possible the value of @xmath103 ( the ratio of @xmath104 is however optimized for the entanglement @xcite ) , which gives a negligible @xmath90 but is large enough to generate sizable entanglement .
another reason @xmath105 must be small is that in order to make the scheme of ref .
@xcite valid @xmath106 must be fulfilled @xcite .
we have also assumed the system is at low temperature @xmath107 mk and pressure @xmath108 torr , resulting in a sufficiently small @xmath91 .
figure [ fig2 ] is plotted as a function of the mechanical frequency @xmath17 which is a key parameter @xcite and can be easily altered by adjusting the intensity of the trapping laser . under these conditions , one could see two distinguishable curves of the total diffusion rate with and without the csl effect ( see fig .
[ fig2 ] ( b ) ) which thereby result in distinguishable curves of the entanglement ( see the appendix for calculating the steady - state entanglement ) , as shown in fig .
[ fig2 ] ( c ) .
we see that the entanglement in the absence of the csl effect decreases almost linearly with @xmath17 , while it increases first and then decreases with the csl effect , and more noticeably , the difference of their values in these two cases is quite large confirming the sensitivity of the entanglement to the noise .
therefore , the csl effect could be determined by repeating the experiment at different trapping laser powers and verifying the distinguishable behavior , particularly the different signs of the slope of the curves , in the region of small @xmath17 . in fig .
[ fig3 ] we present the results for testing the csl with different collapse rates @xmath109 hz , @xmath110 hz , and @xmath111 hz . as @xmath83 decreases , the csl effect becomes weaker and weaker . in order to amplify this effect , we employ bigger and bigger spheres , @xmath5 from @xmath112 to @xmath113 , so that @xmath92 is enlarged while @xmath20 is not increased too much retaining a nonzero entanglement . in this process , @xmath91 is no longer negligible and starts to play its role making the solid lines in fig .
[ fig3 ] ( a)-(c ) less and less linear , especially when @xmath17 is small . since it has the same scaling as @xmath92 versus @xmath17 , i.e. , @xmath91 , @xmath114 ( see table [ table ] ) , its value , depending on the gas temperature and pressure , will be the most relevant factor which eventually determines the value of the upper bound of @xmath83 that can be tested based on the system of levitated nanospheres , either optically or magnetically @xcite trapped .
another efficient way to increase @xmath92 and meantime decrease @xmath20 is by lowering the trapping frequency @xmath17 .
however , the condition @xmath115 limits the smallest value that @xmath17 can take , which however can be relaxed by reducing @xmath18 implemented by taking larger values of @xmath4 .
even so , one should not consider taking too small values of @xmath17 because when the mechanical frequencies are too low other unwanted electronic noises will enter into the system making our scheme less effective .
we see in fig .
[ fig3 ] ( d)-(f ) that , as @xmath83 decreases , the distinguishability of the two curves with and without the csl reduces . at @xmath116
hz , the two curves are no longer that distinguishable ( not like in ( d ) , ( e ) different signs of the slope of the curves for small @xmath17 ) , nevertheless , the difference of their values is still considerable ( thanks to the powerful scheme @xcite ) . if the difference caused by all uncontrolled noises and system errors , e.g. , due to the imprecision of measurements and of the calibration of the experimental parameters , is smaller than that induced by the csl effect ( the relative difference between the two curves in fig .
[ fig3 ] ( c ) , when @xmath6 is small , is about @xmath117 ) , one can determine the csl effect with the collapse rate down to @xmath116 hz , and even lower .
we finally discuss how to detect the generated entanglement of the two nanospheres .
we follow the detection scheme provided in ref .
@xcite , i.e. , sending into the cavity two weak red - detuned probe fields with detunings respectively equal to the two mechanical frequencies , i.e. , @xmath118 ( @xmath119 and @xmath120 are another two cavity resonance frequencies ) .
the probe modes adiabatically follow the dynamics of the two mrs and the output of the readout cavity @xmath121 is given by @xcite @xmath122 where @xmath123 , @xmath124 is the very small optomechanical coupling with the probe mode and @xmath125 denotes the input vacuum noise . therefore , by homodyning the probe mode outputs from the transmission of the controllable beamsplitter ( see fig .
[ fig1 ] ) , and by changing the phases of the corresponding local oscillator , the quadratures of the two mrs , @xmath126 , are measured , and thus the covariance matrix of the quadratures is constructed , from which the entanglement can then be numerically obtained in the way introduced in the appendix .
the probe fields will also induce mechanical diffusion due to the scattering of photons , however , it can be neglected if the probe fields are sufficiently weak .
we have checked that in all plots of fig .
[ fig3 ] , @xmath90 is always negligible compared to other diffusion rates , which means that even if the power of the probe field is equivalent to that of the driving laser , the diffusion induced by it is also negligible .
we have suggested a scheme based on entangled levitating nanospheres to probe the possible effect of the csl theory .
it is designed for nanospheres optically trapped in fabry - prot cavities , however , our scheme can also be applied to the system of magnetically trapped spheres , where the diffusion due to the scattering of trapping field will be significantly reduced @xcite .
we have shown that the steady - state entanglement of the center - of - mass motion of the two nanospheres is particularly sensitive to the noises in the system and the csl effect results in remarkably large reduction of the entanglement .
the entanglement shows distinguishable scalings , with and without the csl effect , with the system parameter of the trapping frequency for the collapse rate down to @xmath127 hz , while it starts to show similar scalings for @xmath128 hz , implying that our scheme can _ unambiguously _ determine the csl effect for @xmath83 down to @xmath129 hz with realistic parameters and can also test it for @xmath128 hz , and even lower , if the difference of the entanglement caused by uncontrolled noises and the imprecision of measurements and parameter calibration is smaller than that induced by the csl effect .
we notice that under the same conditions of the temperature and pressure , @xmath107 mk and @xmath108 torr , the present scheme is not as efficient as the one provided in ref .
@xcite where the observable is instead the phase of the cavity output light .
this is because , as mentioned at the beginning in sec .
[ results ] , in order to have sizable entanglement a relatively small size of the sphere is adopted yielding a larger @xmath91 ( since @xmath99 ) , which eventually determines this scheme , based on entanglement , could not _ unambiguously _ probe the csl effect for @xmath83 down to @xmath0 hz , which however can be done by the scheme of ref . @xcite .
one promising solution is to employ the system of magnetically trapped spheres , e.g. , the proposal provided in ref .
@xcite , where a superconducting microsphere is magnetically trapped close to a quantum circuit .
in contrast with optical levitation , the main diffusion due to the scattering of trapping light is absent and the decoherence in the magnetic levitation system is predicted to be very small .
the new system brings in additional sources of noise , e.g. , due to hysteresis losses in the superconducting coils , fluctuations in the trap frequency and trap center , however , all of them are predicted to be negligible @xcite . adopting such a magnetic levitation scheme , the entanglement could be generated with ( much ) larger size of the spheres , which will significantly increase @xmath92 and meantime reduce @xmath91 . the csl effect with @xmath83 well below @xmath0 hz , if is present , is expected to be probed with realistic parameters .
this work has been supported by the major state basic research development program of china ( grant no .
2012cb921601 ) , the national natural science foundation of china ( grant nos : 11504218 , 11634008 , 11674203 , 91336107 , 61227902 ) .
here we provide the details how we obtain the steady - state entanglement between the two mrs .
the entanglement is calculated based on the covariance matrix of the two mechanical modes .
the covariance matrix can be achieved by solving the quantum langevin equations , which can be rewritten in the following form @xmath130 where @xmath131 is the vector of quadrature fluctuation operators of the two mechanical modes and one cavity mode , i.e. , @xmath132 .
@xmath133 is the drift matrix , which takes the form of @xmath134 where @xmath135 is the effective detuning ( its exact expression is provided in ref .
@xcite ) and we take @xmath136 corresponding to the optimal detuning for the entanglement .
the system is stable when all the eigenvalues of the drift matrix @xmath133 have negative real parts , which can be simply achieved when @xmath137 is fulfilled @xcite .
@xmath138 is the vector of noise quadrature operators associated with the noise terms in the equations .
the _ steady - state _ covariance matrix @xmath139 of the system quadratures , with its entries defined as @xmath140 ( @xmath141 denotes an anticommutator , and @xmath142 ) , is obtained by solving the lyapunov equation @xmath143 where @xmath144 is the diffusion matrix , with its entries defined as @xmath145 the diffusion matrix is a diagonal matrix , which is @xmath146 $ ] .
once the covariance matrix @xmath147 is obtained , the entanglement can then be quantified by means of the logarithmic negativity @xcite : @xmath148,\ ] ] where @xmath149 ( @xmath150 the so - called symplectic matrix and @xmath151 the @xmath152-pauli matrix ) is the minimum symplectic eigenvalue of the covariance matrix @xmath153 , with @xmath154 the @xmath155 covariance matrix associated with the two mechanical modes and @xmath156 the matrix that inverts the sign of momentum of the 2nd mr , i.e. , @xmath157 , realizing partial transposition at the level of covariance matrices @xcite . | wave function collapse models are considered as the modified theories of standard quantum mechanics at the macroscopic level . by introducing nonlinear stochastic terms in the schrdinger equation ,
these models make predictions , differently from those of standard quantum mechanics , that it is fundamentally impossible to prepare macroscopic systems in macroscopic superpositions .
the validity of these models can only be examined by experiments and hence efficient protocols for this kind of experiments are highly needed . here
we provide a protocol that is able to probe the postulated collapse effect by means of the entanglement of the center - of - mass motion of two nanospheres optically trapped in a fabry - prot cavity .
we show that the collapse noise results in large reduction of the steady - state entanglement and the entanglement , with and without the collapse effect , shows distinguishable scalings with certain system parameters , which can be used to unambiguously determine the effect of these models . |
given a large @xmath7 complex matrix @xmath0 and a sufficiently regular function @xmath1 so that @xmath2 is well defined , we are interested in approximating the largest singular values and corresponding singular vectors of the matrix function @xmath2 . this computation will also give an approximation to its 2-norm , namely , @xmath3 , where @xmath8 is the matrix norm induced by the euclidean vector norm , and it is defined as @xmath9 in our presentation we will chiefly discuss this norm approximation because of its interest in applications .
however , we shall keep in mind that the considered procedure allows us to also determine both associated left and right singular vectors , and that a group of singular triplets can be determined simultaneously . the problem of approximating the norm of a matrix function arises in the solution of stiff linear initial value problems @xcite,@xcite , in the evaluation of derivatives and perturbations of matrix functions , which arise for instance in electronic structure theory @xcite,@xcite,@xcite , and in monitoring the magnitude of the inverse of distance matrices @xcite .
in numerical linear algebra the norm of matrix polynomials may be used in the analysis of iterative procedures , and the norm of rational matrix functions , and in particular of the transfer function may give information on the sensitivity of the matrix itself to perturbations ; see , e.g. , @xcite,@xcite and their references . if @xmath0 were normal , then the approximation could be stated in terms of an eigenvalue problem in @xmath0 . indeed , if @xmath10 is the eigendecomposition of @xmath0 with @xmath11 unitary and @xmath12 diagonal , then @xmath13 @xcite , so that the leading singular values of @xmath2 could be determined by a procedure that approximates the eigenvalues of @xmath0 .
the problem is significantly more challenging if @xmath0 is large and non - normal , since there is no relation between eigenvalues and singular values that can be readily exploited during the computation .
moreover , although @xmath0 may be sparse , in general @xmath2 will be dense , and it can not be computed explicitly .
we are thus left with procedures that use @xmath1 and @xmath0 by means of the action of @xmath2 to a vector @xmath14 .
the lanczos bidiagonalization is among the most used strategies for approximating selected singular triplets of a given matrix .
given a matrix @xmath15 , this procedure generates a sequence of orthonormal vectors @xmath16 and @xmath17 by alternating products of @xmath18 and @xmath19 . in our case , @xmath20 and therefore these matrix vector products can be approximately computed .
in fact , since this computation is expensive , we shall consider an _
inexact _ implementation of the lanczos bidiagonalization process , where at each iteration the action of @xmath21 and @xmath22 is approximated with some loose tolerance by means of a projection method .
the problem of approximating @xmath21 has seen a great interest growth in the past fifteen years , due to the emerging occurrence of this computation in many scientific and engineering applications ; see , e.g. , @xcite,@xcite,@xcite,@xcite,@xcite,@xcite,@xcite , and their references . for our purposes
we shall use krylov subspace methods for approximating @xmath22 and @xmath21 at each iteration , equipped with a cheap stopping criterion that may also be adapted to the outer current accuracy .
we shall show that the inexactness in the lanczos bidiagonalization causes the loss of the known symmetry structure of the process .
nonetheless , as is the case in finite precision analysis @xcite , orthogonality of the basis can be preserved , so that the recurrence maintains its effectiveness .
if a rough approximation to @xmath3 is the only quantity of interest , instead of a group of singular triplets , then other approaches could be considered .
for instance , @xmath1 could be approximated by some other more convenient functions , and then the resulting matrix function norm could be more easily estimated . as an alternative ,
equivalent definitions of @xmath2 could be used , from which the norm could also be estimated ; or , the relation of @xmath3 with other norms or with some other spectral tool could be used ; some of these approaches are briefly recalled in section [ sec : known_methods ] .
methods in the mentioned classes , however , usually at most provide the order of magnitude of the actual norm and are thus inappropriate if more correct digits are needed . this paper is organized as follows . section [ sec : known_methods ] reviews some methods available for the approximation of @xmath3 . in section
[ sec : lanczos bidiag ] the standard lanczos bidiagonalization is recalled and the general notation used in this paper is introduced .
section [ sec : inexact lanczos bidiag ] presents the inexact lanczos bidiagonalization procedure , including the details on the stopping criteria in section [ sec : comput_stop ] .
section [ sec : inner ] discusses the approximation of the matrix function multiplication , and a stopping criterion for its accuracy , while in section [ sec : flex ] a stopping criterion in case an inner flexible strategy is analyzed . in section [ sec : spectral properties ] we show how specific spectral properties allow us to make a variable accuracy for the inner iteration feasible , which is finalized in section [ sec : variable ] .
section [ sec : practical implementation ] focuses on the practical implementation and the numerical results are presented in section [ sec : expes ] .
we will conclude with some discussion in section [ sec : discussion ] .
the following notation will be used throughout .
the vector @xmath23 indicates the @xmath24th column of the identity matrix of a given dimension .
the conjugate transpose of a matrix @xmath0 will be denoted by @xmath25 .
we will use the matlab - like notation @xmath26 $ ] to denote the column vector @xmath27 the euclidean vector norm for vectors will be used , namely @xmath28 , for @xmath29 . unless explicitly stated , the induced matrix norm ( [ eqn:2norm ] ) will be used for matrices . for @xmath30 , spec(@xmath0 ) denotes the set of its eigenvalues , and @xmath31 is its field of values .
while the lanczos bidiagonalization is widely recognized as the method of choice for approximating selected singular triplets of a large matrix , if one is only interested in estimates of @xmath32 with @xmath0 non - hermitian , then rather different procedures could also be used .
a simple approach consists of roughly estimating @xmath33 by using some other matrix norm .
for instance , @xmath34 where @xmath35 is once again an induced norm @xcite .
this bound is usually pessimistic , and it is clearly unsatisfactory for @xmath36 large .
the fact that for @xmath0 large the entries of @xmath2 are not all readily available provides an additional challenge . in the following
we describe a few approaches available in the literature that are tailored to the matrix function case .
some of them first determine an explicit upper bound for the norm , which only depends on scalar quantities .
the core computation will then be to determine a good approximation to the obtained upper bound .
the quality of the final estimate of @xmath32 will thus depend both on the sharpness of the initial upper bound and on the accuracy of the computation .
for general non - normal matrices the initial bound is often not very sharp , limiting the quality of the overall estimation .
finally , a computation - oriented estimation is the power method , which directly approximates @xmath32 as the square root of the largest eigenvalue of @xmath37 .
a more detailed list follows . 1 .
let @xmath38 be the numerical radius of @xmath0 , that is @xmath39 , where @xmath40 is the field of values of @xmath0 .
since ( see , e.g. , ( * ? ? ?
* theorem 1.3 - 1 ) ) @xmath41 by applying the bounds to @xmath2 instead of @xmath0 , it is possible to estimate @xmath32 by means of @xmath42 ; see , e.g. , @xcite,@xcite for numerical methods to compute the numerical radius of a given matrix .
a related special case is given by the exponential function , for which the bound @xmath43 holds , where @xmath44 is the largest eigenvalue of the hermitian part of @xmath0 , that is of @xmath45 @xcite .
2 . if it is possible to find @xmath46 and @xmath47 such that @xmath48 then it is sufficient to estimate @xmath49 ; here @xmath49 is the @xmath50-norm of @xmath1 on @xmath51 .
this can be done for instance when @xmath1 is a polynomial , @xcite , for which @xmath52 is known to be less than 11.08 and conjectured to be equal to 2 , and @xmath51 coincides with the field of values of @xmath0 .
we refer to the ph.d .
thesis of d. choi @xcite , for a discussion on the use if this bound when @xmath0 is normal , or when @xmath0 is a contraction ; see also @xcite for a detailed analysis of this bound when using pseudospectral information .
the computational intensive task is given by the determination of @xmath51 .
if @xmath51 coincides with @xmath40 , then the cost of accurately approximating @xmath51 may be superior to that of approximating the single quantity @xmath32 .
this approach is in the same spirit as the one above , but with a higher computational component . for @xmath53 ,
let @xmath54 and assume that @xmath1 is analytic in @xmath55 .
if @xmath56 denotes the length of the boundary @xmath57 , then by using the cauchy integral expression for @xmath2 we obtain ( see , e.g. , @xcite ) @xmath58 although the involved quantities may be easier to compute than in the previous case , the dependence on @xmath53 remains not fully controllable .
4 . using the relation @xmath59 a run of a few iterations of the power method can give an estimate to @xmath60 ; see , e.g. , ( * ? ? ?
* algorithm 3.19 ) for an algorithm specifically designed for the largest singular triplet .
the power method is probably the most appealing approach among the ones listed above .
if a rough approximation is required , typically to determine the order of magnitude , then its low cost provides a satisfactory answer .
however , if more than one digit of accuracy is required , then the process may become slow .
as for with @xmath0 , the stability in the computation may be highly influenced by the squaring ; we refer to section [ sec : expes ] for an example of this well known phenomenon .
we start by recalling the golub - kahan bidiagonalization process in our context , in terms of the matrix function @xmath2 ; then we will discuss how to actually obtain @xmath2 times a vector .
let @xmath61 and @xmath62 , then for @xmath63 the following recurrence relations define the lanczos algorithm @xmath64 the coefficients @xmath65 are computed so that the corresponding vectors @xmath14 s and @xmath66 s have unit norm . by collecting the two sets of vectors
as @xmath67 $ ] , @xmath68 $ ] , we observe that @xmath69 , @xmath70 , @xmath71 . moreover , the two recurrences can be compactly written as @xmath72 where @xmath73 is the following bidiagonal matrix @xmath74 it can be shown that the columns of @xmath75 span the krylov subspace @xmath76 and the columns of @xmath77 span the krylov subspace @xmath78 .
define @xmath79 and @xmath80 then the recursion ( [ eqn : gk ] ) can be rewritten as a standard lanczos process , in the more compact matrix notation ( @xcite , ( * ? ? ?
* , p. 495 ) ) @xmath81 for both even and odd @xmath82 , the eigenvalues of @xmath83 occur in @xmath84 pairs , with the exception of an extra extraneous zero eigenvalue in the odd case . within this setting ,
is it thus possible to approximate the singular values of @xmath85 by the positive eigenvalues of @xmath86 , or , equivalently , by the singular values of @xmath87 . in particular , for the largest singular value it holds that ( see ( * ? ? ?
* corollary @xmath88 , lemma @xmath89 ) ) : @xmath90 there are several advantages of the golub - kahan bidiagonalization over the simpler power method applied to @xmath37 , which are mainly related to the fact that the eigenvalue squaring in this latter problem may lead to severe loss of information in the case very small or very large singular values arise . in the inexact case
the bidiagonal formulation also allows us to better trace the inexactness during the whole approximation process ; this is discussed in the next section .
when neither the explicit computation of the matrix @xmath2 nor the accurate operation @xmath21 ( or @xmath91 ) are feasible , then approximate computations must be performed , resulting in an _ inexact _ lanczos bidiagonalization procedure . as a consequence ,
the recurrence ( [ eqn : gk ] ) needs to be significantly revised so as to aknowledge for the quantities that are actually computed . for a given @xmath14 ,
the exact matrix - vector multiplication @xmath92 has to be replaced by an inner procedure that approximates the resulting vector up to a certain accuracy .
the same holds for the operation @xmath22 for a given vector @xmath66 . for the sake of the analysis , at each iteration
@xmath93 we shall formalize this difference by writing , for some matrices @xmath94 and @xmath95 , @xmath96 where @xmath97 implicitly represent the perturbation induced by the approximate computations .
since in general @xmath98 is no longer the conjugate transpose of @xmath99 , orthogonality of a new vector @xmath100 has to be enforced by explicit orthogonalization with respect to all previous vectors @xmath101 , @xmath102 .
the same holds for the vectors @xmath103 , @xmath104 .
therefore , instead of one bidiagonal matrix @xmath73 in the exact relation , we now obtain an upper triangular matrix @xmath105 and an upper hessenberg matrix @xmath106 .
this leads to the following relations for the inexact ( perturbed ) lanczos bidiagonalization : @xmath107 where @xmath108 and @xmath109 .
the matrices @xmath75 and @xmath77 are different from the matrices in the exact relation , but they still have orthonormal columns
. the inexact lanczos bidiagonalization can also be described using the notation of ( [ eq : lbd with b 2 ] ) .
define @xmath110 , \qquad { \mathcal{w}}_{2 m } = \left [ \begin{array}{cc } u_m & 0 \\ 0 & v_m \end{array } \right],\ ] ] and the perturbation matrix @xmath111 the perturbed relation thus becomes @xmath112{\mathbf e}_{m}^ * , \quad { \mathbf e}_m \in\mathbb{r}^{2m},\ ] ] where @xmath113 in contrast to the exact case , the space spanned by the columns of @xmath114 is not a krylov subspace .
however , when @xmath115 is small , this new space is close to an invariant subspace of the perturbed matrix @xmath116 , because then @xmath117 .
notice the similarity of ( [ eq : inexact lbd k ] ) with equation ( 3.1 ) in @xcite , which shows that with this formulation , the inexact projection problem amounts to solving a structured eigenvalue problem , where the original hermitian matrix @xmath118 has been perturbed by a structured non - hermitian perturbation @xmath119 .
the theory in @xcite can then be used to analyze and monitor the inexact computations , although the general results in @xcite should be carefully adapted to the new problem structure .
if @xmath120 is small in norm , the eigenvalues of the _ non - hermitian _ matrix @xmath121 are small perturbations of the eigenvalues of the _ hermitian _ matrix @xmath118 .
indeed , the eigenvalues of the perturbed matrix @xmath116 lie in discs with radius @xmath122 and center the ( real ) eigenvalues of @xmath118 ( see , e.g. , ( * ? ? ?
* , theorem 5.1 ) ) .
therefore , for small perturbations in the computations , the eigenvalues of the symmetric matrix @xmath118 will be perturbed accordingly . on the other hand , in the following we shall consider the case when @xmath123 is larger than usually allowed by a perturbation analysis argument , therefore different strategies need to be devised to ensure good approximations to the wanted eigenvalues of @xmath118 . following the standard procedure of the exact case , we should consider the matrix @xmath124 to approximate the largest eigenpairs of @xmath116 , and according to the discussion above , of @xmath118 . due to the non - hermitian structure of @xmath124 , however , there are different matrices that can provide the sought after singular value information , namely the matrix @xmath124 itself , and the two distinct matrices @xmath106 or @xmath105 .
the last two matrices yield approximations to the corresponding triplets of @xmath125 and @xmath126 .
the following bound between the largest eigenvalue of @xmath127 and the largest singular values of @xmath106 and @xmath105 shows that all these quantities can be easily related .
let @xmath128 $ ] .
using @xmath129 we obtain . ]
@xmath130\neq 0 } \left|\frac{{\mathbf x}^*m_m{\mathbf y}+ { \mathbf y}^*t_m{\mathbf x}}{{\mathbf x}^*{\mathbf x}+ { \mathbf y}^*{\mathbf y}}\right|\\ & \le & \displaystyle\max_{[{\mathbf x};{\mathbf y}]\neq 0}\frac{\|{\mathbf x}\|\|m_m{\mathbf y}\|}{\|{\mathbf x}\|^2 + \|{\mathbf y}\|^2 } + \max_{[{\mathbf x};{\mathbf y}]\neq 0}\frac{\|{\mathbf y}\|\|t_m{\mathbf x}\|}{\|{\mathbf x}\|^2 + \|{\mathbf y}\|^2}\le
\frac{1}{2}(\sigma_1(m_m ) + \sigma_1(t_m)).\end{aligned}\ ] ] if the inexactness of the bidiagonalization is very large , @xmath105 and @xmath131 are very different from each other . in this case , the leading singular values of these two matrices - and thus their mean - may be significantly larger than the biggest ( in modulo ) eigenvalue of @xmath124 , since they are related to the numerical radius of @xmath124 , rather than to its spectrum .
this motivated us to use the eigenvalues of @xmath124 in the approximation , rather than the singular values of its blocks .
moreover , working with @xmath124 made the analysis of the relaxed strategy particularly convenient , since known results on relaxed eigenvalue computation could be exploited . in this section
we analyze a strategy for monitoring the convergence of the inexact bidiagonal iteration .
as it is common to other inexact processes , the true problem residual is inaccessible as soon as inexactness takes place .
however , on the one hand some stopping criterion needs to be introduced to exit the process . on the other hand ,
it is unclear whether the computed approximations are still meaningful for the original problem , since they were computed with significantly modified data .
let @xmath132 be an eigenpair of @xmath124 , where @xmath133 is a unit vector .
as the iterations proceed , @xmath134 tends to approximate an eigenpair of @xmath127 .
we would like to ensure that @xmath134 also tends to an eigenpair of @xmath118 . to monitor the convergence of @xmath135 and to define a stopping criterion for the outer iteration , the residual is used .
we call @xmath136 the _ true residual _ , which is not available , since @xmath118 can not be applied exactly .
we thus introduce the _ computed residual _ , which is the residual of the actually computed quantities , namely ( see ( [ eq : inexact lbd k ] ) ) @xmath137 in the sequel , we shall use the following obvious inequality to estimate the true residual norm : @xmath138 where @xmath139 is the gap between the computed and the true residuals , in short the `` residual gap '' .
if this gap can be imposed to be small , then the computed residual will give an estimate for the true residual .
in this case , convergence can be monitored by only using the ( available ) computed residual , and the following relative stopping criterion can be used : @xmath140 for some outer tolerance @xmath141 , where @xmath135 is the largest ( in modulo ) eigenvalue . finally , as the computed residual norm goes to zero , the quantity @xmath139 will tend to dominate again , playing the role of the final attainable accuracy level . to see how we can impose the residual gap to be small , and recalling the definition of @xmath142 , we first consider a more convenient expression for @xmath143 , with @xmath144 $ ] , that is @xmath145 let @xmath146 $ ] .
then @xmath147 , @xmath148 , implicitly carry the error caused by the inexact computation of @xmath149 and @xmath150 , respectively , in the inner iteration . if every term of this sum is small , the computed residual will be close to the true residual .
the following lemma states how the inaccuracy in the matrix - vector products relates to the residual gap ; its proof is based on the corresponding result in @xcite , however the structure is exploited so as to have a dependence with respect to @xmath151 instead of @xmath152 , the size of @xmath153 .
[ lemma : res bound ] assume that @xmath151 iterations of the inexact lanczos bidiagonalization process have been taken .
if @xmath154 for @xmath155 , then @xmath156 . _ proof . _ from @xmath157 with @xmath158 $ ] it follows that @xmath159\|\le 1 $ ] . from ( [ eqn : g ] ) we obtain @xmath160 this result shows that if @xmath161 is sufficiently small , then the residual gap will stay below the computed residual norm until convergence . in our experiments
, @xmath151 will play the role of the maximum number of lanczos bidiagonalization iterations , which is usually set to a number between @xmath162 and @xmath163 .
the performance of the inexact lanczos bidiagonalization process depends on the approximation accuracy of the matrix - vector products @xmath149 and @xmath150 . due to the size of @xmath0
, we consider approximating this quantity by means of a projection - type iterative method as follows ; we limit our discussion to @xmath149 , and a corresponding procedure can be used for @xmath150 .
starting with the unit vector @xmath14 and the matrix @xmath0 , we construct a sequence of approximation subspaces @xmath164 of @xmath165 , @xmath166 and define the matrix @xmath167 \in{\mathbb c}^{n\times k}$ ] , whose orthonormal columns span the subspace , and @xmath168 , in a way so that the spaces are nested , that is @xmath169 .
typical such choices are krylov and rational krylov subspaces @xcite,@xcite .
the desired approximation is then obtained as @xmath170 for small @xmath93 , the reduced non - hermitian matrix @xmath171 has small size , so that @xmath172 can be computed efficiently by decomposition - type methods @xcite .
our stopping criterion of this approximation process is based on an estimation of the error norm , and it uses an approach previously introduced in ( * ? ? ?
* proposition 2.2 ) .
* proposition 2.2 ) assume that @xmath173 inner iterations have been executed .
let @xmath174 be an approximation to @xmath149 and define @xmath175 .
if @xmath176 and @xmath177 , then @xmath178 the result in ( [ eqn : error_estimate ] ) shows that after @xmath173 iterations it is possible to provide an estimate of the error norm at iteration @xmath93 .
therefore , we introduce the following stopping criterion for the approximation of @xmath92 : @xmath179 for some inner tolerance @xmath180 . the accuracy of the inner iteration will influence the final accuracy of the inexact lanczos bidiagonalization . in the notation of the previous section ,
once the stopping criterion is satisfied , we have thus derived the following estimate for the perturbation occurring in the lanczos step , @xmath181 note that here @xmath182 , with the notation in ( [ eqn : g ] ) .
an analogous relation holds with respect to @xmath183 and thus @xmath184 .
we stress here that , since the approximation process changes at each iteration , the quantity @xmath185 will vary as the lanczos bidiagonalization proceeds . the threshold itself may vary during the lanczos iteration , so that @xmath186 .
as experienced with other eigenvalue and linear system problems , @xmath187 may even be allowed to grow during the iteration , without significantly affecting the overall process .
this is discussed in the next section .
the bound in ( [ eqn : g ] ) on the residual gap suggests that the accuracy on the inner solution approximation can be relaxed as convergence takes place .
indeed , following similar strategies in @xcite,@xcite,@xcite , we observe that it is the product @xmath188 in ( [ eqn : g ] ) that needs to be small , and not each factor , to ensure a small gap ; the same for @xmath189 .
therefore , if @xmath190 is sufficiently small , indicating that the @xmath191th component of the eigenvector @xmath133 is small , @xmath192 is allowed to be larger , and the required accuracy of @xmath161 can still be achieved .
this induces a variable ( possibly growing ) accuracy in the inner iteration , which drives the size of @xmath192 . in the following
we shall first show that the quantities @xmath190 and @xmath193 do tend to decrease as the approximation improves .
we then derive a computable expression for the variable stopping tolerance in the approximation of @xmath21 and @xmath22 at each iteration of the resulting `` relaxed '' lanczos bidiagonalization process .
this strategy may be convenient in case the cost approximating @xmath21 and @xmath22 is very high , as is the case for instance if an accurate approximation to the leading singular triplets is requested . to ensure that the magnitude of @xmath194 , @xmath195 , can be relaxed in the bound ( [ eqn : g ] ) , we need to verify that @xmath193 and @xmath190 become small as convergence takes place .
this fact has been verified in the eigenvalue setting in @xcite , however the peculiar structure of the lanczos bidiagonal recurrence requires the validation of the results in @xcite . to this end
, we first define the submatrix of @xmath124 of size @xmath196 as @xmath197 where @xmath198 , @xmath199 are the leading portions of the corresponding @xmath200 matrices .
let @xmath201 be an eigenpair of @xmath202 , where @xmath203 $ ] has unit norm , and @xmath204 .
further , let @xmath205 where the @xmath206-vectors have length @xmath207 , and define @xmath208 $ ] , where @xmath209 is chosen such that @xmath210 is unitary .
define @xmath211 .
the following result shows that under certain hypotheses some of the components of the approximate eigenvectors do tend to zero as convergence takes place .
its proof is analogous to that of ( * ? ? ?
2.2 ) , and can be found in the appendix .
[ prop : tau ] let @xmath212 be an eigenpair of @xmath202 , and @xmath213 be as defined in _ ( [ eq : q tilde])_. let @xmath214 , @xmath215 , and @xmath216{\mathbf e}_k^*{\mathbf q}^{(2k)}$ ] . if @xmath217 then there exists a unit norm eigenvector @xmath218 $ ] of @xmath124 with @xmath219 , @xmath220 , such that @xmath221 with @xmath222 , @xmath223 .
moreover , if @xmath135 is the eigenvalue associated with @xmath133 , we have @xmath224 iterations of lanczos bidiagonalization the computed residual @xmath225 is sufficiently small , then there exists an eigenvector of @xmath124 such that some of its components are bounded correspondingly .
these are precisely the components that allow us to relax the accuracy in the inner iteration .
note that @xmath226 gives an indication of the distance between the spectrum of @xmath227 and @xmath228 .
it should be kept in mind that for non - normal matrices , the value of @xmath226 may be much smaller ( * ? ? ?
* example 2.4 , p. 234 ) .
on the other hand , since @xmath229 is a perturbation to a hermitian matrix , the quantity @xmath230 is an approximate residual for @xmath201 as an eigenpair of @xmath229 , and thus it will be small as @xmath151 grows . as a consequence ,
the condition in the theorem is likely to be satisfied , and the eigenvalue error ( [ eqn : thetas ] ) may be much smaller than @xmath231 . in this section
we show that relaxation in the inner accuracy at step @xmath232 is possible if there exists an eigenpair @xmath233 of @xmath234 such that @xmath235 the first condition ( [ restriction 1 ] ) ensures that there exists an eigenvector @xmath133 of @xmath153 whose specified components are small , according to proposition [ prop : tau ] .
let @xmath135 be the eigenvalue associated with this @xmath133 .
the second condition , ( [ restriction 2 ] ) , guarantees that the eigenvalue @xmath236 of @xmath234 is a perturbation of the eigenvalue @xmath135 of @xmath124 , which is the final approximation to the original problem .
the following theorem states how the use of a variable accuracy will still guarantee a small residual gap , and hence yields a true residual with an accuracy which is bounded by the accuracy of the gap , in agreement with ( [ eq : true comp gap ] ) .
assume @xmath151 inexact lanczos bidiagonalization iterations are carried out .
let @xmath132 be an eigenpair of @xmath124 , where @xmath135 is simple and @xmath237 .
given @xmath238 , with the notation of ( [ eqn : g ] ) assume that for @xmath239 @xmath240 then @xmath241 .
_ this proof is analogous to the proof of theorem 3.1 in @xcite .
suppose that at the @xmath242th iteration there exists an eigenpair @xmath233 of @xmath234 satisfying the conditions @xmath243 and @xmath244 .
this implies that @xmath236 is a perturbation of the considered eigenvalue @xmath135 of @xmath124 , since @xmath135 is the only eigenvalue of @xmath124 such that @xmath245 let @xmath246 be defined such that for each @xmath247 there exists a eigenpair @xmath248 of @xmath234 satisfying the conditions @xmath243 and @xmath244 .
then , similar to the reasoning in the proof of lemma [ lemma : res bound ] and using ( [ eqn : g ] ) , @xmath249 [ algorithm1 ] height 0pt depth 0.5pt width + * algorithm 1 : * inexact lanczos bidiagonalization + height 0pt depth 0.3pt width + * input : * @xmath30 non - hermitian , a function @xmath1 , a maximum number of ( outer ) iterations @xmath151 , an ( outer ) tolerance @xmath141 . + * output : * an approximation to the leading singular triplet + height 0pt depth 0.3pt width + [ cols= " > , < " , ]
we have explored the use of an inexact lanczos bidiagonalization method for approximating the leading singular triplet of a large matrix function , and in particular its 2-norm .
although several strategies are known to provide rough estimates of a matrix function 2-norm , more accurate approximations require a careful implementation of available approaches , since neither @xmath2 nor products of the type @xmath21 are available exactly .
in particular , we showed that the lanczos bidiagonalization yields a non - hermitian perturbation of the original hermitian matrix , and the recurrence needs to be revisited .
our numerical experiments showed that the computational complexity may vary significantly depending on the requested final accuracy , since the two inner iterations for the approximation of @xmath21 and @xmath22 may be very time and memory consuming .
we showed that the relaxed strategy alleviates this problem whenever accurate approximations are required .
however , for particular selections of matrices and functions , the approximation of @xmath21 can still be very expensive , and some other strategies could be exploited , such as restarting ; see , e.g. , @xcite,@xcite,@xcite and references therein .
finally , our approach could be used to estimate the norm of other matrix objects , such as the geometric mean @xcite , or the _ derivatives _ of matrix functions , such as the frchet derivative of the matrix exponential or of other functions @xcite . 10 , _ norm estimates for inverses of toeplitz distance matrices _ ,
j. approximation theory , 79 ( 1994 ) , pp .
222242 . ,
_ decay properties of spectral projectors with applications to electronic structure _ , siam review , 55 ( 2013 ) , pp . 364 . , _ norm inequalities related to the matrix geometric mean _ , lin . alg .
appl . , 437 ( 2012 ) , pp .
726733 . , _ inexact matrix - vector products in krylov methods for solving linear systems : a relaxation strategy _ , siam j. matrix analysis and appl .
, 26 ( 2005 ) , pp .
660 678 . , _ estimating norms of matrix functions using numerical ranges _ , phd thesis , university of washington , 2013 .
, _ numerical range and functional calculus in hilbert space _ ,
j. functional analysis , 244 ( 2007 ) , pp .
668690 . , _ lanczos algorithms for large symmetric eigenvalue computations : vol . 1 : theory _ , vol . 41 , siam , 2002 . ,
_ extended krylov subspaces : approximation of the matrix square root and related functions _ , siam j. matrix anal .
appl . , 19 ( 1998 ) , pp . 755771
. , _ solution of large scale evolutionary problems using rational krylov subspaces with optimized shifts _ , tech .
rep . 5 , 2009 . ,
_ a restarted krylov subspace method for the evaluation of matrix functions _ , siam j. numer .
, 44 ( 2006 ) , pp . 24812504 . , _ convergence of restarted krylov subspace methods for stieltjes functions of matrices _ , siam j. matrix anal .
appl , ( 2014 ) . , _ matrix functions _ , tech .
rep . , dipartimento di matematica , bologna , i , march 2006 . to appear on siam j. scient .
computing .
, _ numerical solution of ordinary differential equations : is there anything left to do ? _ , siam rev .
, 23 ( 1981 ) , pp . 1024 . ,
_ perturbation of functions of diagonalizable matrices _ , electronic j. of linear algebra , 20 ( 2010 ) , pp .
303313 . , _ matrix computations _ , the john hopkins university press , baltimore , london , 3rd ed . , 1996 . , _ numerical range : the field of values of linear operators and matrices _ , universitext , springer , new york , ny , usa , 1997 . , _ rational krylov approximation of matrix functions : numerical methods and optimal pole selection _ , gamm mitteilungen , 36 ( 2013 ) , pp
831 . , _ estimating the condition number of the frechet derivative of a matrix function _ , siam j. sci .
comput . , ( 2014 ) . ,
_ matrix functions theory and applications _ , siam , philadelphia , usa , 2008 . , _ exponential integrators for quantum - classical molecular dynamics _ , bit , numerical mathematics , 39 ( 1999 ) , pp . 620645 . , _ exponential integrators _ , acta numerica , 19 ( 2010 ) , pp
. 209286 . ,
_ topics in matrix analysis _ , cambridge university press , cambridge , 1991 . height 2pt depth -1.6pt width 23pt , _ matrix analysis _ , cambridge university press , cambridge , ii ed . , 2013 . ,
_ on the condition number and perturbation of matrix functions for hermitian matrices_. arxiv:1206.1762v1 , june 2012 . ,
_ a new investigation of the extended krylov subspace method for matrix function evaluations _ , numerical linear algebra with applications , 17 ( 2010 ) , pp .
615638 . ,
_ lanczos bidiagonalization with partial reorthogonalization _ , tech .
daimi pb-357 , department of computer science , aarhus university , 1998 . ,
_ a visual repository of test data for use in comparative studies of algorithms for numerical linear algebra , mathematical and computational sciences division , national institute of standards and technology_. online at http://math.nist.gov/matrixmarket . ,
algorithms for the computation of the pseudospectral radius and the numerical radius of a matrix _ , i m a j. num . an . , 25 ( 2005 ) , pp . 648669 . , _ controlling errors in recursive fermi - dirac operator expansions with applications in electronic structure theory _ , siam j. sci .
, 34 ( 2012 ) , pp .
b1b23 . , _ norm bounds for rational matrix functions _ , numerische mathematik , 42 ( 1983 ) , pp .
379389 . , _ variable accuracy of matrix - vector products in projection methods for eigencomputation _ , siam j. numer .
, 43 ( 2005 ) , pp .
11551174 . , _ theory of inexact krylov subspace methods and applications to scientific computing _
, siam j. sci .
, 25 ( 2003 ) , pp .
454477 . ,
_ matrix perturbation theory _ , computer science and scientific computing , academic press inc . , boston , ma , ( 1990 )
, _ spectra and pseudospectra . the behavior of non - normal matrices and operators _ , princeton university press , 2005 . , _ computing the numerical radius _ , lin .
. appl . , 234 ( 1996 ) ,
define the submatrix of @xmath124 of size @xmath196 as @xmath250 define the vector @xmath251 $ ] , where the @xmath206-vectors have length @xmath207 .
let @xmath252 $ ] be such that @xmath210 is unitary , where @xmath253 $ ] .
this implies that @xmath254 , @xmath255 and @xmath256 .
now , write @xmath257 here @xmath258 further , since @xmath259 , we have @xmath260 now , by ( * ? ? ?
* theorem 2.1 , p.230 ) , @xmath261 then there exists a vector @xmath262 satisfying @xmath263 , such that the unit norm vector @xmath264 is an eigenvector of @xmath124 .
moreover , @xmath265 further , this same theorem states that @xmath266 . | given a large square matrix @xmath0 and a sufficiently regular function @xmath1 so that @xmath2 is well defined , we are interested in the approximation of the leading singular values and corresponding singular vectors of @xmath2 , and in particular of @xmath3 , where @xmath4 is the matrix norm induced by the euclidean vector norm .
since neither @xmath2 nor @xmath5 can be computed exactly , we introduce and analyze an _ inexact _ golub - kahan - lanczos bidiagonalization procedure , where the inexactness is related to the inaccuracy of the operations @xmath5 , @xmath6 .
particular outer and inner stopping criteria are devised so as to cope with the lack of a true residual .
numerical experiments with the new algorithm on typical application problems are reported . |
the field of thermoelectricity was born in the 1820s after seebeck reported his observation of a phenomenon he interpreted as thermomagnetism .
orsted , in the same years , gave a better interpretation of the phenomenon and termed it thermoelectricity : the appearance of an electrical current in a circuit subjected to a temperature difference .
the name of seebeck was later given to the coefficient linking the temperature difference @xmath0 and the resulting electromotive force @xmath1 : @xmath2 .
a few years after seebeck s works were published @xcite , peltier noticed that , conversely , an electrical current @xmath3 may absorb or release heat at a junction between two different materials @xcite .
the flux of absorbed heat @xmath4 is related to the electrical current @xmath3 through the difference of peltier coefficients @xmath5 between the materials of the junction : @xmath6 ( with @xmath3 positive when flowing from material @xmath7 to material @xmath8 ) .
the relation between these two effects was later clearly expressed by thomson who established the connection between the peltier and seebeck coefficients : @xmath9 , with @xmath10 the temperature of the considered material @xcite .
this relation demonstrates that both effects have the same physical origin : the ability of charge carriers to carry heat along with electrical charge .
as thermoelectric ( te ) devices possess the particularity to strongly couple heat and electrical transports , they can be used in refrigeration or energy conversion applications @xcite .
however to achieve good performances both te materials and device design have to be optimized . for materials , optimization amounts to increasing the so - called figure of merit @xmath11 ( see for example the recent review by shakouri @xcite ) . yet , to make the best use of the materials it is necessary also to lead a reflection on the integration of the devices into an external environment .
one thus needs a complete model of thermoelectric device to fully comprehend the interaction between the constitutive laws of the device and the laws related to the external exchanges . in this work
we propose a simple model of thermoelectric generator ( teg ) that accounts for the heat transported by electrical carriers ; this results in the introduction of an effective thermal conductance associated with this phenomenon . as this heat propagation
is associated with global movement of particles , by analogy with fluid mechanics , we call it _
convection_. we then study and illustrate the consequences that this additional term may bear on both the physics of te energy conversion and on the practical design and optimization of thermoelectric generators ( tegs ) .
let us consider a thermoelectric generator connected to two heat baths at temperature @xmath12 and @xmath13 ( @xmath14 ) respectively with an average temperature @xmath15 .
though a thermoelectric device may operate indifferently either as a cooler or as a generator , in this work , we restrict our analysis to the case of a generator to which an electrical load resistance @xmath16 is connected in order to extract power .
now that we have set the thermal and electrical environment , we state the phenomenological laws that govern the device by relating the electrical current @xmath3 and the average thermal current @xmath4 to the voltage @xmath1 and the temperature difference @xmath0 across the generator .
the equations follow directly from the force - flux formalism introduced by onsager and then extended to thermoelectricity by callen @xcite . for our purpose
, we find convenient to use the macroscopic coefficients and potential differences , rather than local coefficients and gradients , and we get : @xmath17 @xmath18 where @xmath19 is the electrical resistance , @xmath20 the thermal conductance under open - circuit condition and @xmath21 the seebeck coefficient .
the teg is also characterized by its figure of merit @xmath11 obtained from these three parameters : @xmath22 .
this dimensionless number gives a direct quantitative estimation of the generator s quality @xcite .
the convention for the currents and potential differences , @xmath23 and @xmath24 , are displayed in fig .
( [ fig : figure1 ] ) .
equation ( [ eq : i ] ) gives a definition of the seebeck coefficient under open - circuit condition ( @xmath25 ) , namely @xmath2 .
if the seebeck coefficient @xmath21 vanishes , thermoelectric effects become negligible , and one recovers the traditional macroscopic expressions of ohm s law and fourier s law , from eqs .
( [ eq : i ] ) and ( [ eq : iq ] ) respectively .
the two previous equations may be combined to express the thermal flux as the sum of two terms : @xmath26 and a convective part @xmath27.,scaledwidth=100.0% ] the first term is the contribution due to the heat transported within the electrical current , the so - called _ convective _ thermal flux @xcite , and the second is the conductive part traditionally associated with fourier s law .
the quantity of energy transported by each carrier is @xmath28 , @xmath29 being the elementary electric charge . to compare these two terms we now rewrite eq .
( [ iq1 ] ) to obtain a form similar to fourier s law : a relation of proportionality between the heat current and the temperature difference . for this purpose
we express the electrical current through the resistive load as a function of @xmath0 . from ohm
s law : @xmath30 , we get : @xmath31 after substitution of the above expression for @xmath3 in eq .
( [ iq1 ] ) , the average heat flux may be expressed as : @xmath32 where an additional term , @xmath27 , for the total thermal conductance of the generator now appears clearly .
this term is directly related to the electrical condition on the circuit since the load resistance @xmath33 is explicitly involved in its definition .
the effective thermal conductance of the generator @xmath34 is thereby equivalent to the two thermal conductances @xmath27 and @xmath35 in parallel as illustrated in fig .
( [ fig : figure1 ] ) .
when @xmath36 , i.e. , in open - circuit condition , we recover , as expected , the fact that @xmath34 is equivalent to @xmath35 only , whereas when @xmath37 , @xmath38 . to derive eq .
( [ iq2 ] ) we made a few assumptions : we considered only an average value for the heat flux ; we neglected the joule effect and the thermal energy converted into electric energy along the device .
these approximations are reasonable if the temperature difference @xmath0 is kept small ; a condition that must be satisfied to make use of constant coefficients to describe the te generator .
all these assumptions amount to stating that we only consider a linear framework .
second , we neglected the heat flowing through the electrical load .
so , dealing with a true unileg generator as in ref .
@xcite , implies @xmath39 .
this configuration may also be related to a generator with two legs : the parameters then used are only equivalent ones @xcite , in which case the load sustains no temperature difference , so that no heat flux goes through it .
in this section we illustrate the impact that the convective part of the generator thermal conductance may have on the properties and operation of a teg .
we first look at the construction and meaning of the figure of merit @xmath11 ; next we discuss the question of optimization of thermoelectric generators when the connections to the heat reservoirs are not perfect ; finally we examine the effective thermal conductance of two thermoelectric modules in parallel . to discuss the properties of thermoelectric systems , we find convenient to proceed by analogy using the classical heat transfer description in terms of convection and conduction . to determine whether heat transfer in a fluid is dominated either by conduction or by convection , one may introduce the prandtl number @xmath40 defined as @xmath41 where @xmath42 is the kinematic viscosity of this fluid , characterizing momentum diffusion , and @xmath43 its thermal diffusivity , characterizing heat diffusion @xcite .
we retained this idea to define a _
thermoelectric prandtl number _ @xmath44 given by the ratio of the convective thermal flux and of the conductive thermal flux : @xmath45 it appears that @xmath44 is proportional @xmath11 .
the proportionality coefficient depends on the ratio between the electrical resistances of the generator and the load .
if the electrical circuit is shorted , i.e. , if @xmath46 , the thermoelectric prandtl number @xmath44 is then equal to @xmath11 , which is the maximum value of the ratio between the convective part and the conductive part of the thermal flux .
it is consistent with the fact that @xmath11 must be increased to improve the teg performances : the convective thermal flux is a direct consequence of the coupling between heat and electrical charge transport , and , as such , it is a necessary process to obtain energy conversion .
conversely , the conductive thermal flux amounts to pure loss effects only : so maximizing @xmath11 is equivalent to maximizing the _ useful part _ of the thermal flux with respect to the losses .
the ratio @xmath47 controls the convective component of the heat flux .
however , maximization of @xmath44 does not lead to the best performances : to extract power from the heat that flows in the generator , one has to satisfy impedance matching too .
if @xmath33 becomes too small , the electrical current , and hence the convective thermal current , will increase but the power will decrease .
it appears that the optimal value of @xmath44 associated to the maximum power output is not @xmath11 but @xmath48 .
we can even extend this analysis to the maximization of the teg efficiency : as shown by ioffe @xcite , for a teg with constant parameters the efficiency is maximized when @xmath49 .
thereby , for the efficiency the optimal value of @xmath44 is @xmath50 .
this last result should be compared with the compatibility approach of thermoelectric optimization : snyder and ursell @xcite demonstrated that in order to obtain the best efficiency , the relative current @xmath51 ( with @xmath52 the electrical current density and @xmath53 the local thermal conductivity under open circuit - condition ) has to be kept constant along the device and should be equal to an optimal value @xmath54 . expressed differently
the condition for efficiency maximization reads : @xmath55 . as the term on
the left hand side is a local definition of @xmath44 , we recover exactly the same result as above .
it is interesting to note that similar conditions were derived using the ratio of the total heat current to the convective thermal flux in refs .
. finally the notion of convective thermal conductance can also be useful to practically determine the figure of merit of a thermoelectric generator : in 1966 , lisker proposed to obtain @xmath11 by measuring the effective thermal conductance of the te module for two different electrical load conditions instead of extracting from three different measurements the value of each parameter @xmath19 , @xmath21 and @xmath20 @xcite .
the ratio @xmath56 is indeed equal to @xmath57 .
assume that the generator is not directly coupled to the heat reservoirs ; the finite thermal conductance of the contacts must be taken into account .
these are denoted @xmath58 on the hot side and @xmath59 on the cold side of the generator .
the total equivalent conductance for both contacts is given by @xmath60 .
the main effect of these imperfect contacts is that the actual temperature difference across the teg is not @xmath0 but @xmath61 as illustrated on the left side of fig .
( [ fig : figure2 ] ) .
we first need to determine the relation between these two temperature differences .
to do so we use an analogue to the potential divider formula for the temperature , assuming that the heat flux remains constant along the structure ( this hypothesis is the same as that used to derive the expression of @xmath27 ) .
we obtain the following relation : @xmath62 since the thermal conductance @xmath34 depends on the electrical load resistance through @xmath27 , the temperature difference across the generator , @xmath63 , is modulated by the electrical load condition .
the consequence of this modulation is that the electrical circuit modeling of the teg as diplayed in fig .
( [ fig : figure2 ] ) is no longer a thvenin generator : in such model the voltage source must be independent of the load which is not the case when finite thermal contacts are introduced . to keep using a proper thvenin generator we split the thermoelectric voltage @xmath64 into two terms :
one is independent of the electrical load , and the other is linked to an internal resistance @xcite : @xmath65 this equation is of the form : @xmath66 with @xmath67 .
we thus obtain a rigorous thvenin modeling of the electrical part of the teg with the definitions of the open circuit voltage given by @xmath68 and the internal resistance is @xmath69 : as for the effective thermal conductance , the effective electrical resistance of the generator shows an additional contribution directly related to the thermal conditions imposed on the generator through the thermoelectric coupling coefficient @xmath21 .
further , it is remarkable that the expressions of @xmath70 and @xmath27 present the same form .
one can switch from one to the other by changing thermal conductances to electrical resistances and _ vice versa_. recently , the introduction of an extra electrical resistance depending on thermal conditions was also suggested by spry @xcite . with our model , the power produced by the teg may thus be simply expressed as : @xmath71 from which we deduce that power maximization is obtained for @xmath72 .
such a condition was already derived mathematically by freunek and co - workers @xcite . for different values of zt ( right).,scaledwidth=100.0% ] as regards thermal matching , stevens proposed that to maximize power production , the teg should be thermally matched with @xmath73 .
let us examine the situation using the equations we just derived .
an expression for the maximum power may be obtained using eq .
( [ pprod1 ] ) in the case of electrical matching and the definition of @xmath74 : @xmath75 for a given value of @xmath11 , @xmath76 shows a clear dependence on the ratio @xmath77 . if we consider a fixed value of @xmath78 , due for example to technological constraints , the choice of the value for @xmath79 will strongly impact the ability of the generator to produce power . on the right side of fig .
( [ fig : figure2 ] ) we show the relation between @xmath76 and @xmath77 for three different values of @xmath11 ( 1 , 3 and 10 ) . note that for each point electrical impedance matching is achieved .
we see that this condition for power maximization through thermal impedance matching slightly differs from the one obtained by stevens , and that it depends on @xmath11 .
actually the optimum value of @xmath20 corresponds to the thermal matching between @xmath78 and @xmath34 : @xmath80 @xcite .
the dependence on @xmath11 is explained by the fact that the greater @xmath11 is , the greater the contribution of @xmath27 to @xmath34 is .
thus for sufficiently high teg efficiency , i.e. , high @xmath11 , the discrepancy between these two conditions of thermal impedance matching is not negligible : see the difference in fig .
( [ fig : figure2 ] ) between the maximum power and the value of power for the condition @xmath81 displayed by the vertical dot - dashed line .
finally we express the simultaneous conditions of thermal and electrical matching in terms of ratios @xmath82 and @xmath47 ( as @xmath19 and @xmath35 are more suitable quantities for device design than @xmath34 and @xmath74 ) : @xmath83 @xmath84 once again , we emphasize the symmetry that arises between thermal and electrical quantities . in a recent publication , yazawa and
shakouri obtained the same relations from a more mathematical viewpoint @xcite .
our last example concerns the determination of the effective thermal conductance of two thermoelectric modules in parallel as represented in fig .
( [ fig : figure3 ] ) .
we keep the same notations for modules properties except that we add a subscript ( @xmath85 or @xmath86 ) to distinguish the two modules .
first , we derive the total average thermal flux @xmath87 inside the system : it is given by the sum of the average thermal flux for each module , with a conductive and a convective part for each .
so we have : @xmath88 as we focus on the open electrical circuit condition to determine the equivalent thermal conductance @xmath89 , the total electrical current is set to @xmath25 .
however this does not mean that the currents @xmath90 and @xmath91 vanish as well : the only consequence is that @xmath92 .
a simple analysis of the electrical circuit in fig . ( [ fig : figure3 ] ) leads to the following expression for the internal current : @xmath93 this internal current only exists if the electromotive forces , given by @xmath94 , do not compensate each other , i.e. , if @xmath95 . in that case we have to consider the convective contribution to the thermal flux . substituting eq .
( [ eq : i1 ] ) in eq .
( [ eq : conservationiq ] ) and simplifying by @xmath0 we get the equivalent open circuit thermal conductance : @xmath96 this last equation demonstrates that the convective component of the thermal flux has a significant impact on the effective thermal conductance of the two thermoelectric modules in parallel if they possess very dissimilar seebeck coefficients .
however this term is of limited interest if the electrical resistances of the modules are too high : the magnitude of electrical internal current is then small and , consequently , so is the convective thermal flux .
knowledge of the effective thermal conductance may be useful when dealing with superlattices for which the layers can be viewed as parallel modules @xcite .
more surprisingly , price obtained a similar expression for the effective thermal conductance of an ambipolar conductor where the subscripts 1 and 2 are associated with electrons and holes repectively : the two types of carriers are indeed considered as evolving in parallel inside the system @xcite .
convective thermal flux will have more and more impact on the behavior of thermoelectric generator as the performance of materials will be increased .
indeed we have shown that the proportion of convective to conductive thermal current is directly related to the figure of merit @xmath11 which is expected to increase significantly in the future to allow a wider use of thermoelectric power production . in this paper
we have demonstrated that this additional contribution to heat flow should not be neglected for device design as well as for materials research , especially those concerning composites where internal currents can develop and lead to a performance decrease .
this work is part of the syspacte projects funded by the fonds unifi interministriel 7 .
y. a. acknowledges financial support from the ministre de lenseignement suprieur et de la recherche . | coupling between heat and electrical currents is at the heart of thermoelectric processes . in a thermoelectric system
this may be seen , from a thermal viewpoint , as an additional thermal flux linked to the appearance of an electrical current . since this additional flux
is associated with the global displacement of charge carriers in the system , it can be qualified as convective in opposition to the conductive part related to both phonon transport and heat transport by electrons under open circuit condition as , e.g. , in the wiedemann - franz relation . in this article we demonstrate that considering the convective part of the thermal flux allows both new insight into the thermoelectric energy conversion and the derivation of the maximum power condition for generators with realistic thermal coupling . |
most of the gravitational properties of cosmic strings @xcite and superconducting cosmic strings @xcite have so far been calculated in the special framework of zero @xcite or time independent currents @xcite so that it is widely believed that observation of strings in the universe might be achieved through the now well known double image effect , i.e. the deflection associated with the missing angle in the conical metric surrounding a string .
an interesting observation would therefore consist in two exactly identical stars ( or galaxies ) separated by a fixed angle ( usually assumed of the order @xmath1rad
@xmath2 , i.e. , for strings generated at the grand unified phase transition ) , or perhaps an aligned series of such twin images @xcite .
however , because the probability that two nearly identical stars be separated in the sky by pure coincidence in exactly the way one expects for cosmic strings is rather high , there is in fact a demand for a more convincing signature .
as we shall see , a lightlike current may in principle provide such a signature . in this paper , we study the gravitational field created by a lightlike current - carrying string . for the case of timelike and spacelike currents , which have been studied before @xcite ,
all known solutions are stationary @xcite . in the lightlike case ,
however , we shall see that the stationary solution has peculiar asymptotic behaviour , in the sense that all light rays are gravitationally bound by the string .
therefore , the problem of asymptotic light deflection is not well posed in this metric .
this is just as well , since in realistic situations , we do not expect the current to be stationary . in particular we can consider a pulse of lightlike current travelling along the string at the speed of light .
such pulses may be generated when a long string interacts with a bounded external source . in this case , the motion of the double image with time due to the nonstationary gravitational field must be examined .
the article is organized as follows : in a first section , we derive the metric surrounding a lighlike current carrying cosmic string by boosting a spacelike current metric , and show how the resulting gravitational field can be generalized to a time - dependent exact solution of the einstein - maxwell equations . in section [ wit ]
we address the question of finding the corresponding solutions in the underlying field theoretic model . for definiteness , we use the bosonic witten model @xcite coupled to gravity .
we show that the introduction of a lightlike current does not modify the vortex profile .
also , the effect of the current on the metric is accounted for by means of a single function which satisfies a linear differential equation . in order to understand the effect of a lightlike current on a possible observation of light deflection by a string , we use the fact that our solution can in fact account for time - varying fields , so we consider a finite pulse of such a current .
we then work out analytically the characteristic features of the geodesic motion in this metric , and eventually calculate numerically , in the last section , the actual effective motion ( meaning as seen by a geodesic observer ) of a point source initially aligned with a string and an observer when a pulse of lightlike current passes between the source and the observer .
we conclude by considering the astrophysical observational possibilities and discussing a possible mechanism for generating pulses of lightlike current in cosmic strings .
in this section , we shall obtain the gravitational field surrounding a cosmic string carrying a lightlike current , treating the source as an infinitely thin distribution .
one easy way to do so is by the limiting procedure @xcite of boosting the metric of the spacelike current case which is known @xcite to the speed of light in the direction parallel to the string ( alternatively , one could also boost the timelike current metric , both procedures being equivalent ) . in the case of spacelike current , for a static cylindrically symmetric configuration with the string lying along the @xmath3axis ,
the metric outside the core of the string is @xcite @xmath4 here , @xmath5 ^ 2 $ ] and @xmath6 , where @xmath7 is the thickness of the core of the string . the nonvanishing component of the electromagnetic field is given by @xmath8 where @xmath9 is newton s constant .
the various parameters @xmath10 , @xmath11 , @xmath12 and @xmath13 should be determined , in principle , in terms of the microphysical parameters characterizing the vortex by matching the exterior metric ( [ metricini ] ) with the solution of einstein s equations inside the core of the string ( i.e. , for @xmath14 ) . in practice
, this can be difficult if the gravitational field is strong .
however , for weak fields , one can arrive at the relations @xcite @xmath15+{\cal o}(g),\nonumber\\ \null & & \label{match } \\
c_2 & = & { 1\over 2}[1-{g^{1/2}\over i}(u - t - i^2)]+{\cal o}(g),\nonumber\\ \gamma & = & 1 - 4g(u+{1\over 2}i^2)+{\cal o}(g^2).\nonumber\end{aligned}\ ] ] here , @xmath16 , @xmath17 and @xmath18 are , respectively , the flat space values of the string s energy per unit length , tension and current ( see , e.g. , refs .
@xcite for a discussion of the limiting procedure consisting in taking well - defined flat space integrals of the stress energy tensor and current as the source of the weak gravitational field ) . treating the string as an idealized distribution of zero thickness
, the components of the string s energy momentum tensor are given by @xmath19 , [ @xmath20 $ ] , where @xcite @xmath21 a boost in the @xmath22 direction @xmath23 transforms eq .
( [ metricini ] ) into @xmath24 where @xmath25 , @xmath26 and we have dropped the prime on the coordinates .
also the current , which was purely spacelike @xmath27 , [ @xmath28 , transforms into @xmath29 , and becomes null as @xmath30 .
the @xmath31 components of the stress tensor ( [ taumunu ] ) transform into @xmath32 @xmath33 , aquiring non diagonal pieces .
it is clear that if we want the components of @xmath34 and @xmath35 to be finite , the limit @xmath36 can only be taken by letting @xmath18 and @xmath37 go to zero at the same time , in such a way that the products @xmath38 remain finite . upon so doing , we find @xmath39 where we have used null coordinates @xmath40 and @xmath41 .
using eqs .
( [ match ] ) and ( [ finite ] ) , the function @xmath42 can be written as @xmath43 also we can write @xmath44 from eq .
( [ elecini ] ) , the electromagnetic tensor in the new frame is @xmath45 and the energy momentum of the string is given by @xmath46 where we have used eqs .
( [ tauab ] ) and ( [ finite ] ) .
typically @xcite , the degeneracy @xmath47 is of order @xmath48 , where @xmath49 is the electromagnetic coupling .
therefore , from eq .
( [ finite ] ) @xmath50 in cosmological applications , @xmath51 . since @xmath52 ,
both terms in eq.([potential ] ) will be of the same order of magnitude .
although the metric ( [ metfin],[potential ] ) has been obtained using the linearized relations ( [ match ] ) , we shall see that it is actually a solution of the full einstein - maxwell equations , with electromagnetic field given by eq .
( [ newf ] ) .
this fact , which is also encountered in the aichelburgh
sexl case @xcite , has a simple mathematical explanation . as @xmath36 ,
we let @xmath18 and @xmath37 to zero , and so we are boosting a metric for which ( in the limit @xmath36 ) the linear approximation becomes exact . for later use , it is convenient to generalize eq .
( [ metfin ] ) to the case where , instead of having a stationary null current flowing along the string , we have a localized pulse of current which is travelling in the positive @xmath22 direction at the speed of light
. the metric ( [ metfin ] ) belongs to the class known as @xmath53waves @xcite @xmath54 for the metric ( [ pp ] ) , the only nonvanishing component of the ricci tensor outside the core is @xmath55 the only component of the electromagnetic energy momentum tensor is @xmath56 taking the ansatz @xmath57 with @xmath42 given by eq .
( [ potential ] ) , it is clear that the einstein equation ( [ e ] ) is satisfied outside the core . in the core of the string , our metric ( [ pp ] )
should be matched with an interior solution . instead of that
, we shall treat the source as infinitely thin , with support only at @xmath58 . in a space with a deficit angle @xmath59
, we have @xmath60 where @xmath61 and @xmath62 is the laplacian in the ( @xmath63 ) plane .
then , one can see that the source corresponding to ( [ pp]-[hur ] ) has the form @xmath64 the first term , corresponding to the goto - nambu string , is responsible for the deficit angle .
the second term , caused by the nondegeneracy @xmath65 , is modulated by the function @xmath66 that gives the profile of the pulse of current . from @xmath67
, we have @xmath68 if @xmath66 has compact support , the metric ( [ pp ] ) represents the gravitational field of a pulse of current travelling in the positive @xmath22 direction .
in the previous section , we treated the superconducting string as an infinitely thin distribution along which a pulse of lightlike current flows . the question remains , however , of whether it is possible to find the corresponding solutions in the underlying field theoretic model coupled to gravity .
for definiteness , we shall consider the witten bosonic model @xcite , consisting of two complex scalar fields and associated u(1 ) gauge fields ( @xmath69 ) and ( @xmath70 ) .
the lagrangian is given by @xmath71 here , @xmath72 , @xmath73 , @xmath74 and @xmath75 .
the potential @xmath76 undergoes spontaneous symmetry breaking .
since the vacuum manifold is nontrivial , @xmath77 will admit vortex ( string - like ) solutions .
@xmath78 develops a condensate in the core of the string , but vanishes ouside .
the field @xmath79 is then identified with electromagnetism , which is unbroken outside the string . to find the gravitational field of a superconducting cosmic string with lightlike current in the model ( [ lagwit ] )
, we shall proceed in two steps .
first , we shall consider the metric for the static and cylindrically symmetric configuration in which no current is flowing in the string .
this configuration is also invariant with respect to boosts parallel to the symmetry axis @xcite .
then , we shall see that the metric for the current - carrying case is just a taub - kerr - schild generalization of the previous one . in the case
where there is no current , we can make the following ansatz for the fields and the metric @xcite @xmath80 @xmath81 the only nonvanishing component of @xmath82 is taken to be @xmath83 and , for the time being , we set @xmath84 and @xmath85const
. the equations of motion for the fields then reduce to @xmath86 these have to be supplemented with the einstein s equations for @xmath87 and @xmath88 .
the energy - momentum tensor s nonvanishing components are @xmath89\label{trr}\ ] ] @xmath90 -{1\over 2 } e^b \left ( \varphi ' ^2 + \sigma ' ^2 + 2 { \cal v}\right ) \label{tthth}\ ] ] @xmath91 where @xmath92 is the lagrangian @xmath93 - { \cal v } .\label{lr}\ ] ] einstein s equations then take the form @xcite @xmath94 actually , the third of eqs .
( [ 18 ] ) is not independent of the other two @xcite
. therefore , eqs .
( [ fieldmot ] ) and ( [ 18 ] ) form a system of five coupled differential equations for the five unknown @xmath95 , @xmath96 , @xmath97 , @xmath98 and @xmath99 , with boundary conditions @xmath100 , @xmath101 , @xmath102 , @xmath103 , @xmath104 , @xmath105 , @xmath106 , @xmath107 and @xmath108 as @xmath109
. these equations are far too complicated to solve analytically , even in flat space ( @xmath110 ) , where they have been studied only numerically @xcite .
however , it is generally believed that such solutions should exist , and that , at least for @xmath111 , their asymptotic properties can be inferred from what is known to be true in the flat space case @xcite . in this case , the energy density is concentrated within a region of radius @xmath112 and , for @xmath113 , the energy density falls off exponentially .
therefore , when gravity is considered , we expect that the metric should behave as a vacuum solution for @xmath114 .
static cylindrically symmetric vacuum solutions fall into the class given by eq .
( [ metricini ] ) , with @xmath115 . if , in addition , we demand boost invariance along the axis , we must take @xmath116 or @xmath117 . for @xmath117 ,
the metric is of kasner type and has the property that a circle around the axis at @xmath118 has zero length .
such metric is considered unphysical from the point of view of strings . for @xmath116 ,
the exterior metric is just flat space with a deficit angle which is given by @xcite @xmath119,\label{18 * 1}\ ] ] where @xmath120 the only difference with the nielsen - olesen case is that we now have the condensate @xmath97 in the core of the string , which will modify the value of @xmath121 [ see eq .
( [ tuv ] ) ] , and hence @xmath122 , but the metric is still basically flat space minus a small wedge . now , to introduce a lightlike current on the string , we take the ansatz ( [ vortex ] ) and ( [ btheta ] ) for @xmath77 , @xmath78 and @xmath82 , but we introduce a nonvanishing gauge field @xmath79 and @xmath123-dependent phase @xmath124 .
@xmath125 as we shall see below , this is consistent with an ansatz for the metric of the form @xmath126 the equation for @xmath79 following from ( [ lagwit ] ) is @xmath127 where @xmath128 is the electromagnetic current .
( [ 21 ] ) and ( [ apsi ] ) with the metric ( [ 20 ] ) require @xmath129 where @xmath130 is a constant , and @xmath131 it is interesting to observe that the introduction of the fields ( [ apsi ] ) and the new term @xmath132 in the metric do not affect the form of eqs .
( [ fieldmot ] ) .
as we shall see , the equations for @xmath98 and @xmath99 are also unaffected and therefore , the null current does not change at all the profile of the vortex ( unlike the case of spacelike or timelike currents ) . to write down einstein s equations
, we use the fact that , as mentionned before , the metric ( [ 20 ] ) is the taub - kerr - schild generalisation of ( [ metexp ] ) . a direct application of taub s formalism @xcite ( see , e.g. eq.(3.23 ) of ref .
@xcite ) yields the new ricci tensor @xmath133 where @xmath134 is the old ricci tensor , given in terms of @xmath98 and @xmath99 by eqs .
( [ 18 ] ) , and @xmath135 also , it is easily verified that the new energy - momentum tensor can be written as @xmath136 where @xmath137 is the old energy - momentum tensor , given by eqs .
( [ trr]-[tuv ] ) , and the new component is given by @xmath138 -e^a h { \cal l},\label{25}.\ ] ] the trace @xmath139 is unaffected , and so the einstein equations ( [ 18 ] ) are unchanged , as promised .
however , a new equation , corresponding to the @xmath140 component , has to be considered : @xmath141 = 8\pi g \widehat t_{uu } .\label{26}\ ] ] using ( [ 18 ] ) , we have @xmath142 hence , from ( [ 25 ] ) and ( [ 26 ] ) , one arrives at the simple linear equation for @xmath143 @xmath144.\label{27}\ ] ] the general solution to this can be obtained as the sum of a particular solution plus the general solution of the homogeous equation @xmath145 to find the particular solution , we take the ansatz @xmath146 which implies @xmath147,\label{29}\ ] ] where @xmath148 is the laplacian in the transverse plane .
note that ( [ 29 ] ) is just an ordinary differential equation .
note also that outside the core , @xmath149 , @xmath150 and eq .
( [ 27 ] ) reduces to ( [ e ] ) of the previous section .
also , if the thickness of the core is small with respect to all other relevant length scales , then it is justified to replace the term @xmath151 in the right hand side of ( [ 27 ] ) by a delta function distribution . to summarize the results of this section
, we can say that the lightlike current actually decouples from the background fields forming the static string , since the introduction of the null gauge field has no effect on @xmath97 , @xmath96 , @xmath98 and @xmath99 .
its only effect on the space time geometry can be accounted for by solving the linear ordinary differential equation ( [ 29 ] ) . as mentionned before , to the particular solution given by ( [ 29 ] )
, we can add any solution of the homogeneous equation ( [ 28 ] ) , @xmath152 .
such solutions can accomodate an additional gravitational wave , as well as a cosmic string transverse wave travelling in the same direction as the null current @xcite .
in the stationary case , i.e. , when the metric takes the form ( [ metfin ] ) , one can find three constants of motion for geodesic test particles @xmath153 where @xmath154 [ @xmath122 can be expressed in terms of underlying fields through ( [ 18 * 2 ] ) ] . in ( [ 30 ] ) , a dot denotes derivative with respect to an affine parameter @xmath155 . for convenience ,
we take @xmath156 , where @xmath157 is the proper time , and @xmath12 is the mass of the particle . using the constraint @xmath158 , we have @xmath159 and the problem reduces to the motion of a newtonian particle in a potential @xmath42 given by ( [ potential ] ) . note
, however , that the stationary metric has peculiar asymptotic properties .
in particular , since the potential grows logarithmically with radius , the motion in the @xmath160 direction will be bounded even for light rays ( @xmath161 ) .
therefore , the problem of asymptotic light deflection is not well posed because we do not have asymptotic regions where light rays would propagate along straight lines .
of course , this extreme behaviour will not arise if , instead of the stationary metric , we consider a finite pulse of current , because then the metric is flat before and after the passage of the wave . for simplicity
, we can take a pulse with a step function profile .
the metric is given by ( [ pp ] ) with @xmath162 where @xmath98 is the duration of the pulse ( not to be confused with the metric function in the previous section ) .
the geodesic equation for the @xmath163 coordinate reads @xmath164 .\ ] ] therefore , @xmath165 defined in ( [ 30 ] ) is only piecewise constant in this case , undergoing jumps when the particle enters and exits the wave .
denoting by @xmath166 the coordinate radius where the wave hits the particle ( at @xmath167 ) and by @xmath168 the radius at which the particle emerges from the wave ( at @xmath169 ) , we have @xmath170 where @xmath171 and @xmath172 and @xmath173 are the values of @xmath174 before and after the particle has interacted with the wave . without loss of generality , we take @xmath175 as an initial condition .
if @xmath176 initially , we can always perform a lorentz transformation to a new frame where such component of the velocity will vanish [ under this lorentz transformation , @xmath177 and @xmath178 , where @xmath179 . then , in the new frame , @xmath180 , and , from ( [ 33 ] ) , we find that the velocity of the particle parallel to the string ( after the wave has past ) is @xmath181,\label{45b}\ ] ] ( typically , @xmath182 ) . hence , the particle receives a boost in the @xmath22 direction whose sign and magnitude are completely determined by @xmath166 and @xmath168 . in particular , if the particle is initially at rest , it will be attracted towards the string by the passing wave , so that @xmath183 [ i.e. , @xmath184 and the boost will be in the positive @xmath22 direction .
on the other hand , a light ray in the @xmath185 plane will be deflected in the @xmath22 direction by the wave , but the sign of the deflection depends on whether the interaction with the wave happens when the ray is approaching the string or when the ray has already surpassed the string and is moving away from it . in the former case , the deflection is in the positive @xmath22 direction , whereas in the latter it is in the negative @xmath22 direction . in the next section , we shall numerically treat the motion of null geodesics in the spacetime given by ( [ 32 ] ) , but for the rest of this section , we shall concentrate on a special case which can be treated analytically , namely the shock wave case @xcite @xmath186 this can be considered as limiting case of ( [ 32 ] ) when the thickness of the wave is much smaller than other relevant dimensions in the problem ( such as the impact parameter or the distance of the wave and of the observer to the string )
. it will be convenient to use cartesian coordinates in which the metric ( [ pp ] ) reads @xmath187 where @xmath188 , @xmath189 , with @xmath190 , so that there is a wedge of angle @xmath191 missing in the @xmath192 plane ( see fig .
[ fig : wedge ] ) .
we shall take the source to be sufficiently far away so that the incident rays are parallel to the @xmath193 axis .
the geodesic equations read @xmath194 and @xmath195 can be obtained from the constraint @xmath196 using ( [ 34 ] ) , we have @xmath197 here , @xmath198 is the initial velocity , @xmath199 , @xmath200 and @xmath201 are the coordinate where the particle hits the wave , and @xmath202 . for light rays lying in the @xmath192 plane , we have @xmath203 . from ( [ 37 ] ) , @xmath204 with @xmath205 and @xmath206 given by ( [ 38 ] ) . denoting by @xmath207 the deflection in the @xmath192 plane and
@xmath96 the deflection in the @xmath22 direction , we have , to first order in @xmath208 , @xmath209 @xmath210 where the last term in the r.h.s .
of ( [ 40 ] ) corresponds to the deficit angle created by the static string .
an interesting question , from the observational point of view , is to find out what is the effect of the wave on the double image caused by the static string .
as we shall see , the two images undergo apparent motion , describing ( almost ) closed trajectories in the sky . for simplicity , we shall take @xmath211 where @xmath212 , with @xmath213 a cosmological scale .
this is a good approximation to ( [ potential ] ) , since in the course of the scattering , @xmath214 changes only by a small percentage . from the geometry of fig .
[ fig : wedge ] , it is clear that the deflection angle is given by @xmath215 where @xmath216 is the distance of the observer to the string . comparing with ( [ 40 ] ) , we have @xmath217 this equation gives the locus of points at which the rays received by the observer have been scattered by the wave . from it , we can obtain @xmath218 , and , substituting in ( [ 39 ] ) and ( [ 40 ] ) , @xmath200 can be eliminated to obtain the curve @xmath219 . since ( [ 41 ] ) is a complicated expression , it is difficult to carry out this procedure in general .
however , we can easily obtain the answer in two limiting cases .
the first limiting case is when @xmath220 in that case , we can replace @xmath201 by @xmath221 in ( [ 39 ] ) and ( [ 40 ] ) to find @xmath222 ^2 \approx \left [ { a \alpha \over 16g\mu \ell \pi}\right ] ^2 , \label{42}\ ] ] that is , the double images will describe circles of angular radii @xmath223 $ ] .
the second limiting case which can be solved is when the deficit angle can be neglected , @xmath224 . in that case , we have @xmath225 the image describes the following trajectory .
initially , it moves upwards to an angular distance @xmath226 from its initial position without moving sideways .
then it splits into two images which describe semicircles of radius @xmath226 .
finally , the two images merge at the bottom of the circle and eventually go back to the original position .
however , for all its beauty , this pattern will not arise in situations involving realistic cosmic strings . typically , the effect of the deficit angle is always comparable , if not much larger , than the effect due to the current .
the reason is that , although a cosmic string can in principle support an arbitrarily large lightlike current , it is difficult to think of a mechanism to generate currents much stronger than the `` critical current '' , defined by the condition @xmath227 .
we shall return to this question in the concluding section .
equations ( [ 42 ] ) and ( [ seclim ] ) correspond to the case of a shock wave . if the duration of the wave is @xmath228 max@xmath229 $ ] , then the shock wave approximation is no longer a good description , since the duration is large compared with the impact parameter @xmath99 .
in such a case , one can see from eqs .
( [ 45b ] ) and ( [ 36 ] ) that to first order in @xmath230 the deflection angles are of order @xmath231 and @xmath232 .
in the previous section we considered the apparent motion of double images in the sky due to the light deflection caused by an impulsive wave . however , we neglected the fact that due to the gravitational field , the observer and the source will undergo geodesic motion , which will also contribute to the apparent motion . here , we shall include this effect , studying the actual motion of a point source as seen from the observer . also , we shall consider a wave of finite duration @xmath98 .
we shall assume that initially the source and the observer are both in a plane orthogonal to the string , and that neither of them is at infinite distance .
we are therefore looking at the particular configuration shown in fig .
[ fig : not ] , where the angles of observation @xmath233 and @xmath234 are defined as the usual azimuthal and polar angles .
it turns out that it will be simpler to use , instead of the forms ( [ 30 ] ) and ( [ 31 ] ) for the geodesic equations where the constant of the motions are explicitely exhibited , the full second order differential equations directly derived from the metric ( [ pp ] ) with @xmath235 given by ( [ hur ] ) : @xmath236 @xmath237 @xmath238 here a dot still means differentiation with respect to the affine parameter @xmath155 along the geodesic under consideration and we have explicitely made use of the conservation of @xmath239 .
we consider the case where the null current is of constant amplitude over a compact support , i.e. , the function @xmath235 is given by ( [ 32 ] ) . now , there is a problem with the set of coordinates we have been working with to derive eqs .
( [ ddotr ] ) , ( [ ddotth ] ) and ( [ ddotz ] ) . indeed ,
as the wave reaches the observer , the latter will experience motion in the @xmath22 direction [ see e.g. , eq .
( [ obsloc ] ) below ] . to avoid this unphysical `` frame dragging '' effect
, it is convenient to pick a particular coordinate system in which the observer and the source are at rest . assuming a free - falling observer of mass @xmath240 for which we neglect the motion in the radial direction , we have , according to eqs.([pp ] ) @xmath241 and since eqs .
( [ 30 ] ) tell us that @xmath123 is proportional to the proper time @xmath242 , we are free to choose @xmath243 so that eq .
( [ obs ] ) transforms into @xmath244 incidentally , this equation also tells us that @xmath245 so the motion in the @xmath22 direction after the wave has past is @xmath246 where @xmath216 is the distance from the observer to the string .
we see that setting @xmath247 ( @xmath123 being unchanged in the new frame ) implies @xmath248 which vanishes at the observer s and at the source s locations according to eq .
( [ obsloc ] ) .
dropping the tildes , we now have the following equation to solve numerically : @xmath249 where @xmath250 while eqs . ( [ ddotr ] ) and ( [ ddotth ] ) still hold . that the approximation of neglecting radial motion is valid can be checked by considering the actual radial displacement experienced by the observer . according to eq .
( [ ddotr ] ) with the previous choice @xmath243 , and assuming the observer to be initially at rest , one has @xmath251 where we neglected variations of @xmath42 with @xmath123 , since this effect is of higher order .
thus , at most , the correction in @xmath160 reads @xmath252 with @xmath253 as discussed earlier .
therefore , since in practical application , one expects @xmath254 , the motion in the radial direction can be confidently neglected .
we have investigated numerically the solutions of eqs .
( [ ddotr ] ) , ( [ ddotth ] ) and ( [ newtz ] ) for various values of the parameters describing the pulse , and the results are displayed on figs .
[ fig : pht ] , [ fig : tht ] and [ fig : phitheff ] . on theses figures ,
we have assumed a gut string of core radius @xmath255mpc , and the internal string parameters have been enhanced to @xmath256 and @xmath257 in order to magnify the corresponding lensing effect .
the string is located at @xmath258mpc from the observer and @xmath259mpc from the source ( see fig . [
fig : not ] ) , and the duration of the pulse is @xmath260mpc .
this particular example provides an illustration of a generic situation .
we are mainly interested by what the observer actually sees , that is , with the notation of fig .
[ fig : not ] , we want to know the variations of the angles @xmath233 and @xmath234 with time , i.e. , we calculate the values of @xmath261 and @xmath262 as a function of the time of arrival of each light ray .
the predicted observations are the following .
initially , the observer sees the usual double image ( the two parallel straight lines on fig .
[ fig : tht ] ) until the wave reaches the first deflected ray .
then , the two images appear to move away from each other , while in the same time moving upwards ( fig . [
fig : pht ] ) . after a time characteristic of the pulse duration ( or of the impact parameter
if this is larger than the pulse duration ) , the images begin to go down and eventually towards each other , and finally they reach a new position in the sky as the wave has passed . put together , figs .
[ fig : pht ] and [ fig : tht ] yield fig .
[ fig : phitheff ] which shows the actual apparent trajectory of the source .
the source is seen to follow two open symmetric curves , i.e. , the final location of the source in the sky is not the same as the initial one .
this can be understood as follow : as the wave passes through the observer and the source , they both reach new positions in space , but because they were not initially at the same distance to the string ( only case for which we expect closed curves at this level of approximation ) , they have not travelled the same distances in the @xmath22 direction , the variation in @xmath22 being @xmath263 as discussed earlier . assuming the source to be farther than the observer implies that the latter has travelled more than the former which is thus seen below its initial location .
more precisely , the actual angle of observation will be @xmath264 , with @xmath230 as defined previously on eq .
( [ alpha ] ) .
in addition to this effect there is a frame rotation effect : since the observer moves in both radial and @xmath22 directions , its frame is being rotated ( parallel transported ) , so that he will see the source at an angle @xmath265 because of this rotation . in practice , @xmath266 will be negligible due to the large denominators , and the curves will be almost closed . another possible way to measure the effect of
the wave is through the redshift of the source : since each light ray experiences a different gravitational potential , one would expect to observe the source with a frequency that should vary while the wave passes . for
the purpose of calculating this effect , let us define the redshift @xmath267 by the relation @xmath268 here @xmath269 is the four - velocity of the observer or of the source , whereas @xmath270 is the four - momentum of the light ray .
we have @xmath271 .
[ fig : red ] shows this redshift as a function of the time of arrival of the light rays with , for comparison , the corresponding angles @xmath233 and @xmath272 , and assuming the same parameters ( @xmath273 , @xmath216 , @xmath98 , @xmath16 , @xmath17 , @xmath18 and @xmath7 ) as on previous figures .
the redshift ( solid line ) is seen to be strongly correlated with the deflection in the @xmath234 direction .
it should be clear that simultaneous measurements of effective motion of double images ( figs .
[ fig : pht ] to [ fig : phitheff ] ) together with the reshift variability shown on fig .
[ fig : red ] would provide an unambiguous signature of the existence of cosmic strings endowed with superconducting currents .
the characteristic time scale during which the source is actually moving in the sky can be evaluated as roughly the wave duration if this one is large enough ( see fig . [
fig : a ] ) .
however , one may expect astrophysical situations in which @xmath98 is much smaller than any other length involved in the problem .
in particular , for the idealized shock wave case of eq .
( [ 34 ] ) , the observation time should be of the order of the impact parameter @xmath274 . in practice ,
the actual time during which the motion of the source takes place is thus @xmath275,\ ] ] which , in the case of a gut string located at a few mpc from us , would be of a few years .
this can be seen more precisely on fig .
[ fig : a].@xmath99 where the time variability of the angle @xmath234 is shown for various values of the wave duration @xmath98 : on these curves , it is clear that the characteristic time in which the signal reaches its maximum is essentially independent of @xmath98 , and in fact reflects only the distance of the string to the observer . finally , let us ask the question of the dependence of these results with the various geometric parameters @xmath273 , @xmath216 and @xmath98 .
as long as @xmath276 , the results do not depend strongly on @xmath273 and we can take @xmath277 as we did in the previous section . for fixed @xmath273 and @xmath98 , the effect decreases with increasing @xmath216 . also displayed on fig .
[ fig : phitheff ] is the previously discussed result for @xmath278 and 100 mpc , with a fixed value @xmath260 mpc of the wave duration .
the effect decreases with increasing @xmath216 , although only by a factor of order 1 .
since we are in the case in which the duration of the wave is much longer than the impact parameter @xmath279 , the angular deflection is of order @xmath230 in the @xmath233 direction and of order @xmath280 in the @xmath234 direction , as discussed at the end of the last section .
if the source is much closer to the string than the observer is , i.e. , in the limit @xmath281 the angular deflection would be suppressed by a geometrical factor of order @xmath282 .
we have derived an exact solution of the einstein s equations that describes the gravitational field surrounding a lightlike current carrying cosmic string .
the current can be of arbitrary shape and time dependence . in the thin string limit
, the exterior metric belongs to the general class of @xmath0-wave solutions .
this exterior metric can be matched with the solution of the witten model coupled to gravity in the core of the string .
the resulting metric is just a taub - kerr - shild generalization of the metric of a cosmic string with vanishing current .
an interesting feature of the solution is that both the function @xmath283 characterizing the electromagnetic field and the function @xmath284 describing the gravitational disturbance obey linear equations [ see eqs .
( [ 22 ] ) and ( [ 27 ] ) ] .
this fact can be understood heuristically as a consequence of the lorentz invariance of the `` seed '' solution and the lightlike nature of the perturbation . indeed , as mentioned before , under a lorentz transformation of velocity @xmath285 parallel to the string we have @xmath286 and @xmath287 , where @xmath288 .
physically , this reflects the fact that the source is redshifted for @xmath289 .
therefore , by making @xmath285 sufficiently close to 1 , we can make @xmath143 and @xmath290 as small as we wish .
thus , it is not surprising that these quantities should obey linear equations .
this also explains why the profile of the vortex and the metric functions @xmath87 and @xmath88 of section [ wit ] are unchanged by introducing the lightlike current .
we calculated the motion of light rays for the case of a pulse current of finite duration @xmath98 and constant amplitude characterized by the parameter @xmath230 given in eq .
( [ alpha ] ) .
in addition , we have the parameter @xmath291 characterizing the mass scale of the cosmic string ( typically we have @xmath292 for gut strings ) , and the geometric parameters @xmath273 and @xmath216 , which give the distance from the string to the source and the observer respectively .
we restricted attention to the case when source and observer are colineal with the string and are initially in a plane orthogonal to the string axis , obtaining the following results .
the initial image is the usual double image before the lightlike current passes , and each one of the double images describes an ( almost ) closed loop in the sky as the wave passes . for @xmath276 , the angular size of the loop is of order @xmath230 in the direction perpendicular to the string and of order @xmath293 in the direction parallel to the string . here
@xmath279 is the typical impact parameter and we are assuming @xmath294 .
we have also seen that the images experience a blueshift and a redshift which is strongly correlated with the apparent motion in the direction parallel to the string , and which is also of order @xmath293 . for sources very close to the string , such that @xmath295 , the magnitude of the angular deflection is decreased by a geometrical factor roughly of order @xmath282 , although for obvious reasons , the redshift effect is unaffected by this factor .
these results only apply if the duration of the pulse is large compared with the impact parameter , that is , if @xmath296 max @xmath297 $ ] . in the opposite limit , the pulse can be approximated by a delta function shock wave . sending the source to infinite distance
, the apparent motion of the double images can be calculated analytically . when the effect of the deficit angle is dominant with respect to the effect of the passing wave ( more precisely , when @xmath298 ) , the double images describe circular motion in the sky , of angular radius given by @xmath299 .
the time scale needed for the image to go around the curve reflects the impact parameter as well as the duration of the pulse .
the effect grows to an angular size of order @xmath230 in a characteristic time of order @xmath300 . if @xmath301 the effect persist for a time of order @xmath98 , and then decreases again in a time scale of order @xmath300 .
for instance , a gut string endowed with a short pulse of gut lightlike current located at a few mpc from us , would deflect an infinitely far source in a few years time , and along a curve of angular aperture of the order @xmath1rad , a phenomenon which should be within the reach of observational detection limits .
in fact , an observation of this kind requires quite particular conditions due to the scarcity in the number of strings that we expect out to such distances @xcite . nonetheless , we believe that the low observational probability is largely compensated by the uniqueness of the signal : such an observation would provide a clear proof of the existence of cosmic strings , and show that they are of the superconducting current - carrying kind .
let us estimate this observation probability more precisely by proposing a mechanism through which lightlike currents can build in cosmic string , namely the interaction of such a string with a galaxy , or any region having an electromagnetic field .
in the original witten model @xcite , the currents built in such a way are small compared with the gut scale . however , in models with spontaneous current generation @xcite one may be able to build up saturation currents in such a way .
if this interaction region is finite , and if the string can be approximated as straight across it , then a current will be induced @xcite , which can be represented by means of a phase function @xmath302 along the string as @xmath303 .
now , away from the interaction region , where the external electromagnetic field vanishes , the general solution for @xmath302 ( to first order in @xmath49 ) is @xcite @xmath304 and must be constant if no interaction takes place .
so , far from the interaction region , one has two waves of null current , one travelling in the positive @xmath22 direction and the other in the negative @xmath22 direction , both at the speed of light , and with a duration of the order of the size of the interaction region . as mentioned before
, the null current does not affect the profile of the string , and so there is in principle no limit to the amount of lightlike current that a string can carry .
however , since in the interaction region the current will be typically spacelike or timelike , and such currents achieve saturation , it is clear that we can not build up lightlike currents whose strength @xmath230 is much larger than @xmath291 .
we see therefore that a requirement for observing a string s lightlike current pulse is that there exist string loops crossing galaxies say , and whose size would exceed that of the interaction region .
now we know that the string loop distribution is such that @xcite @xmath305 for the number @xmath306 of loops of size @xmath307 located in a sphere of radius @xmath216 . here , @xmath308 is the age of universe ( 10@xmath309 yr ) and @xmath310 is a parameter obtained by numerical simulations .
the coherence length of a string due to wiggles is @xmath311 , where @xmath312 @xcite and so we want @xmath307 to be at least this value .
moreover , the impact parameter @xmath279 being the typical duration of the effective motion of the source , we require it to be @xmath313 yr so as to be able to actually measure the effect .
this means that we require @xmath314 yr @xmath315 .
thus , the expected number of events can be estimated as @xmath316 which can be quite small for gut strings , but otherwise increases tremendously for lower mass strings , and is of order one for @xmath317 .
this shows that lighter strings would perhaps produce a detectable ( at the level of arc seconds ) double image moving according to our calculations with a correspondingly varying redshift .
it should be mentioned that this work has been involved with a lightlike current wave separating two regions of the string where there was no current .
lightlike currents may also exist at the boundary between spacelike and timelike currents .
the time evolution of such lightlike currents and their gravitational effect is left for further research .
we should also say that for cosmic strings without currents , time dependent exact solutions have been studied in the past , e.g. in refs .
@xcite . in these cases ,
the positions of the double images are also time dependent , although the actual apparent motion is qualitatively different from our case .
finally , one may consider solutions describing the collision of two pulses of lightlike current travelling in opposite directions . before the collision the metric is a trivial superposition of two solutions like the one we have considered in this paper , but during and after the collision the solution will contain interesting non - linear effects . in particular
, since the metric in the impulsive case resembles very much the aichelburg - sexl metric @xcite ( representing the gravitational field of an ultrarelativistic body ) , and it is known that the head on collision of two such shock waves generates trapped surfaces @xcite , it would be interesting to study the possibility that the collision of such lightlike currents would produce black holes .
we are grateful to peter death , allen everett , joao magueijo , paul shellard , takahiro tanaka and tanmay vachaspati for suggestions and conversations during the preparation of this work .
this project was supported by serc grant # 15091-aoz - l9 .
t. m. helliwell , d. a. konkowski , , 1918 ( 1986 ) ; i. moss , s. poletti , phys .
b * 199 * , 34 ( 1987 ) ; a. babul , t. piran , d. n. spergel , _ ibid _ * 209 * , 477 ( 1988 ) ; p. amsterdamski , p. laguna - castillo , , 877 ( 1988 ) ; b. linet , class .
quantum grav .
* 6 * , 435 ( 1989 ) p. aichelburgh , r. sexl , g. r. g. * 2 * , 303 ( 1971 ) ; t. dray , g. t hooft , nucl .
b * 253 * , 173 ( 1985 ) ; r. jackiw , d. kabat , m. ortiz , phys . lett .
b * 277 * , 148 ( 1992 ) ; c. o. loust , n. sanchez , phys .
b * 232 * , 462 ( 1989 ) e. copeland , m. hindmarsh , n. turok , phys .
58 * , 1910 ( 1987 ) ; r. l. davis , e. p. s. shellard , phys .
b * 209 * , 485 ( 1988 ) ; _ ibid _ * 207 * , 404 ( 1988 ) ; c. t. hill , h. m. hodges , m. s. turner , , 263 ( 1988 ) ; a. babul , t. piran , d. n. spergel , phys .
b * 202 * , 307 ( 1988 ) ; e. copeland , d. haws , m. hindmarsh , n. turok , nucl .
b * 306 * , 908 ( 1988 ) ; d. haws , m. hindmarsh , n. turok , phys .
b * 209 * , 255 ( 1988 ) p. peter , , 1091 ( 1992 ) ; _ ibid _ * 46 * , 3335 ( 1992 ) ; _ ibid _ * 47 * , 3169 ( 1993 ) d. kramer _ et al .
_ , `` exact solutions of einstein field equations '' , v. e. b. deutscher verlag , berlin 1980 ; c. klimik , phys .
b * 208 * , 373 ( 1988 ) ; j. garriga , e. verdaguer , , 391 ( 1991 ) ; d. amati , c. klimik , phys . lett .
b * 210 * , 92 ( 1988 ) ; _ ibid _ * 219 * , 443 ( 1989 ) ; g. t. horowitz , a. steif , phys .
lett . * 64 * , 260 ( 1990 ) ; g. w. gibbons , comm .
* 45 * , 191 ( 1975 ) | the lensing properties of superconducting cosmic strings endowed with a time dependent pulse of lightlike current are investigated .
the metric outside the core of the string belongs to the @xmath0wave class , with a deficit angle .
we study the field theoretic bosonic witten model coupled to gravity , and we show that the full metric ( both outside and inside the core ) is a taub - kerr - shild generalization of that for the static string with no current .
it is shown that the double image due to the deficit angle evolves in an unambiguous way as a pulse of lightlike current passes between the source and the observer .
observational consequences of this signature of the existence of cosmic strings are briefly discussed . |
what do reported uncertainties actually tell us about the accuracy of scientific measurements and the likelihood that different measurements will disagree ?
no scientist expects different research studies to always agree , but the frequent failure of published research to be confirmed has generated much concern about scientific reproducibility @xcite .
when scientists investigate many quantities in very large amounts of data , interesting but ultimately false results may occur by chance and are often published . in particle physics ,
bitter experience with frequent failures to confirm such results eventually led to an ad hoc `` 5-sigma '' discovery criterion @xcite , i.e. a `` discovery '' is only taken seriously if the estimated probability for observing the result without new physics is less than the chance of a single sample from a normal distribution being more than five standard deviations ( `` @xmath1 '' ) from the mean . in other fields , arguments that most novel discoveries are false @xcite
have caused increased emphasis on reporting the value and uncertainty of measured quantities , not just whether the value is statistically different from zero @xcite .
research confirmation is then judged by how well independent studies agree according to their reported uncertainties , so assessing reproducibility requires accurate evaluation and realistic understanding of these uncertainties .
this understanding is also required when analysing data , combining studies in meta - analyses , or making scientific , business , or policy judgments based on research .
the experience of research fields such as physics , where values and uncertainties have long been regularly reported , may provide some guidance on what reproducibility can reasonably be expected @xcite . most recent investigations into reproducibility focus on
how often observed effects disappear in subsequent research , revealing strong selection bias in published results .
removing such bias is extremely important , but may not reduce the absolute number of false discoveries since not publishing non - significant results does not make the `` discoveries '' go away . controlling the rate of false discoveries depends on establishing criteria that reflect real measurement uncertainties , especially the likelihood of extreme fluctuations and outliers @xcite .
outliers are observations that disagree by an abnormal amount with other measurements of the same quantity . despite every scientist knowing that the rate of outliers is always greater than naively expected , there is no widely accepted heuristic for estimating the size or shape of these long tails .
these estimates are often assumed to be approximately normal ( gaussian ) , but it is easy to find examples where this is clearly untrue @xcite . to examine the accuracy of reported uncertainties ,
this paper reviews multiple published measurements of many different quantities , looking at the differences between measurements of each quantity normalized by their reported uncertainties .
previous similar studies @xcite reported on only a few hundred to a few thousand measurements , mostly in subatomic physics .
this study reports on how well multiple measurements of the same quantity agree , and hence what are reasonable expectations for the reproducibility of published scientific measurements .
of particular interest is the frequency of large disagreements which usually reflect unexpected systematic effects .
sources of uncertainty are often categorized as statistical or systematic , and their methods of evaluation classified as type a or b @xcite .
type a evaluations are based on observed frequency distributions ; type b evaluations use other methods .
statistical uncertainties are always evaluated from primary data using type a methods , and can in principle be made arbitrarily small by repeated measurement or large enough sample size .
uncertainties due to systematic effects may be evaluated by either type a or b methods , and fall into several overlapping classes @xcite .
class 1 systematics , which include many calibration and background uncertainties , are evaluated by type a methods using ancillary data .
class 2 systematics are almost everything else that might bias a measurement , and are caused by a lack of knowledge or uncertainty in the measurement model , such as the reading error of an instrument or the uncertainties in monte carlo estimates of corrections to the measurement .
class 3 systematics are theoretical uncertainties in the interpretation of a measurement . for example , determining the proton radius using the lamb shift of muonic hydrogen requires over 20 theoretical corrections @xcite that are potential sources of uncertainty in the proton radius , even if the actual measurement of the lamb shift is perfect .
the uncertainties associated with class 2 and 3 systematic effects can not be made arbitrarily small by simply getting more data . when considering the likelihood of extreme fluctuations in measurements , mistakes and `` unknown unknowns '' are particularly important , but they are usually assumed to be statistically intractable and are not often considered in traditional uncertainty analysis .
mistakes are `` unknown knowns '' , i.e. something that is thought to be known but is not , and it is believed that good scientists should not make mistakes .
`` unknown unknowns '' are factors that affect a measurement but are unknown and unanticipated based on past experience and knowledge @xcite . for example , during the first 5 years of operation of lep ( the large electron positron collider ) , the effect of local railway traffic on measurements of the @xmath2 boson mass was an `` unknown unknown '' that no - one thought about .
then improved monitoring revealed unexpected variations in the accelerator magnetic field , and after much investigation these variations were found to be caused by electric rail line ground leakage currents flowing through the lep vacuum pipe @xcite .
in general , systematic effects are challenging to , but can be partially constrained by researchers making multiple internal and external consistency checks : is the result compatible with previous data or theoretical expectations ?
is the same result obtained for different times , places , assumptions , instruments , or subgroups ?
as described by dorsey @xcite , scientists `` change every condition that seems by any chance likely to affect the result , and some that do not , in every case pushing the change well beyond any that seems at all likely '' .
if an inconsistency is observed and its cause understood , the problem can often be fixed and new data taken , or the effect monitored and corrections made . if the cause can not be identified , however , then the observed dispersion of values must be included in the uncertainty .
the existence of unknown systematic effects or mistakes may be revealed by consistency checks @xcite , but small unknown systematics and mistakes are unlikely to be noticed if they do not affect the measurement by more than the expected uncertainty .
even large problems can be missed by chance ( see sec . [
ssec : modelling ] ) or if the conditions changed between consistency checks do not alter the size of the systematic effect .
the power of consistency checks is limited by the impossibility of completely changing all apparatus , methods , theory , and researchers between measurements , so one can never be certain that all significant systematic effects have been identified .
quantities were only included in this study if they are significant enough to have warranted at least five independent measurements with clearly stated uncertainties .
medical and health research data were extracted from some of the many meta - analyses published by the cochrane collaboration @xcite ; a total of 5580 measurements of 310 quantities generating 99433 comparison pairs were included . particle physics data ( 8469 measurements , 864 quantities , 53988 pairs ) were retrieved from the review of particle physics @xcite . nuclear physics data ( 12380 measurements , 1437 quantities , 66677 pairs )
were obtained from the table of radionuclides @xcite .
most nuclear and particle physics measurements have prior experimental or theoretical expectations which may influence results from nominally independent experiments @xcite , and medical research has similar biases @xcite , so this study also includes a large sample of interlaboratory studies that do not have precise prior expectations for their results . in these studies , multiple independent laboratories
measure the same quantity and compare results .
for example , the same mass standard might be measured by national laboratories in different countries , or an unknown archaeological sample might be divided and distributed to many labs , with each lab reporting back its carbon-14 measurement of the sample s age .
none of the laboratories knows the expected value for the quantity nor the results from other labs , so there should be no expectation , selection , or publication biases .
these interlab studies ( 14097 measurements , 617 quantities , 965416 pairs ) were selected from a wide range of sources in fields such as analytical chemistry , environmental sciences , metrology , and toxicology .
the measurements ranged from genetic contamination of food to high precision comparison of fundamental physical standards , and were carried out by a mix of national , university , and commercial laboratories . all quantities analysed are listed in the supplementary materials @xcite .
data were entered using a variety of semi - automatic scripts , optical - character recognition , and manual methods .
no attempt was made to recalculate past results based on current knowledge , or to remove results that were later retracted or amended , since the original paper was the best result at the time it was published .
when the review of particle physics @xcite noted that earlier data had been dropped , the missing results were retrieved from previous editions @xcite . to ensure that measurements were as independent as possible , measurements were excluded if they were obviously not independent of other data already included . because relationships between measurements are often obscure , however , there undoubtedly remain many correlations between the published results used .
medical and health data were selected from the 8105 reviews in the cochrane database @xcite as of 25 september 2013 .
data were analysed from 221 intervention reviews whose abstract mentioned @xmath3 trials with @xmath4 total participants , and which reported at least one analysis with @xmath5 studies with @xmath6 total participants . the average heterogeneity inconsistency index ( @xmath7 ) @xcite is about 40% for the analyses reported here .
because analyses within a review may be correlated , only a maximum of 3 analyses and 5 comparison groups were included from any one review .
about 80% of the cochrane results are the ratio of intervention and control binomial probabilities , e.g. mortality rates for a drug and a placebo .
such ratios are not normal @xcite , so they were converted to differences that should be normal in the gaussian limit , i.e. when the group size @xmath8 and probability @xmath9 are such that @xmath8 , @xmath10 , and @xmath11 are all @xmath12 , so the binomial distribution converges towards a gaussian distribution .
( the median observed values for these data were @xmath13 , @xmath14 . )
the 68.3% binomial probability confidence interval was calculated for both the intervention and control groups to determine the uncertainties .
measurements with uncertainties are typically reported as @xmath15 , which means that the interval @xmath16 to @xmath17 contains with some defined probability `` the values that could reasonably be attributed to the measurand '' @xcite .
most frequently , uncertainty intervals are given as @xmath18 , where @xmath19 is the coverage factor and @xmath20 is the `` standard uncertainty '' , i.e. the uncertainty of a measurement expressed as the standard deviation of the expected dispersion of values .
uncertainties in particle physics and medicine are often instead reported as the bounds of either 68.3% or 95% confidence intervals , which for a normal distribution are equivalent to the @xmath21 and @xmath22 standard uncertainty intervals . for this study , all uncertainties were converted to nominal 68.3% confidence interval uncertainties .
the vast majority of measurements reported simple single uncertainties , but if more than a single uncertainty was reported , e.g. `` statistical '' and `` systematic '' , they were added in quadrature . all measurements , @xmath23 , of a given quantity were combined in all possible pairs and the difference between the two measurements of each pair calculated in units of their combined uncertainty @xmath24 : @xmath25 the dispersion of @xmath26 values can be used to judge whether independent measurements of a quantity are `` compatible '' @xcite .
a feature of @xmath27 as a metric for measurement agreement is that it does not require a reference value for the quantity .
( the challenges and effects of using reference values are discussed in section [ ssec : alternate ] . ) the uncertainties in equation [ eq : zdefinition ] are combined in quadrature , as expected for standard uncertainties of independent measurements .
( the effects of any lack of independence are discussed in section [ ssec : uncertaintyschemes ] . )
uncertainties based on confidence intervals may not be symmetric about the reported value , which is the case for about 13% of particle , 6% of medical , 0.3% of nuclear , and 0.06% of interlab measurements .
following common ( albeit imperfect ) practice @xcite , if the reported plus and minus uncertainties were asymmetric , @xmath26 was calculated from eq .
[ eq : zdefinition ] using the uncertainty for the side towards the other member of the comparison pair .
for example , if @xmath28 , @xmath29 , and @xmath30 , then @xmath31 and @xmath32 .
the distributions of the @xmath26 differences are histogrammed in fig .
[ fig : differential ] , with each pair weighted such that the total weight for a quantity is the number of measurements of that quantity .
for example , if a quantity has 10 measurements , there are 45 possible pairs , and each entry has a weight of 10/45 .
( other weighting schemes are discussed in section [ ssec : alternate ] . )
the final frequency distribution within each research area is then normalized so that its total observed probability adds up to 1 . if the measurement uncertainties are well evaluated and correspond to normally distributed probabilities for @xmath33 , then @xmath27 is expected to be normally distributed with a standard deviation @xmath34 .
probability distribution uncertainties ( e.g. the vertical error bars in fig . [
fig : differential ] ) were evaluated using a bootstrap monte carlo method where quantities were drawn randomly with replacement from the actual data set until the number of monte carlo quantities equaled the actual number of quantities .
the resulting artificial data set was histogrammed , the process repeated 1000 times , and the standard deviations of the monte carlo probabilities calculated for each @xmath27 bin . random selection of measurements instead of quantities was not chosen for uncertainty evaluation because of the corrections then required to avoid bias and artifacts .
for example , if measurements are randomly drawn , a quantity with only 5 measurements will often be missing from the artificial data set for having too few ( @xmath35 ) measurements drawn , or if it does have 5 measurements some of them will be duplicates generating unrealistic @xmath36 values , or if duplicates are excluded then they will always be the same 5 nonrandom measurements . without correcting for such effects , the resulting measurement monte carlo generated uncertainties are too small to be consistent with the observed bin - to - bin fluctuations in fig . [
fig : differential ] . correcting for such effects
requires using characteristics of the actual quantities and would be effectively equivalent to using random quantities .
attempts were made to fit the data to a wide variety of functions , but by far the best fits were to non - standardized student - t probability density distributions with @xmath37 degrees of freedom . @xmath38
a student - t distribution is essentially a smoothly symmetric normalizable power - law , with @xmath39 for @xmath40 .
the fitted parameter @xmath41 defines the core width and overall scale of the distribution and is equal to the standard deviation in the gaussian limit and to the half - width at half maximum in the cauchy ( also known as lorentzian or breit - wigner ) limit .
the parameter @xmath37 determines the size of the tails , with small @xmath37 corresponding to large tails .
the values and standard uncertainties in @xmath41 and @xmath37 were determined from a non - linear least squares fit to the data that minimizes the nominal @xmath42 @xcite : @xmath43 where @xmath44 , @xmath45 and @xmath46 are the bin @xmath27 , contents , and uncertainties of the observed @xmath27 distributions shown in fig .
[ fig : differential ] .
possible values of @xmath27 are sometimes limited by the allowed range of measurement values , which could suppress heavy tails .
for example , many quantities are fractions that must lie between 0 and 1 , and there is less room for two measurements with 10% uncertainty to disagree by @xmath0 than for two 0.01% measurements .
the size of this effect was estimated using monte carlo methods to generate simulated data based on the values and uncertainties of the actual data , constrained by any obvious bounds on their allowed values .
the simulated data was then fit to see if applying the bounds changed the fitted values for @xmath41 and @xmath37 .
the largest effect was for medical data where @xmath37 was reduced by about 0.1 when minimally restrictive bounds were assumed .
stronger bounds might exist for some quantities , but determining them would require careful measurement - by - measurement assessment beyond the scope of this study .
for example , each measurement of the long - term duration of the effect of a medical drug or treatment would have an upper bound set by the length of that study .
since correcting for bounds can only make @xmath37 smaller ( corresponding to even heavier tails ) , and the observed effects were negligible , no corrections were applied to the values of @xmath37 reported here .
histograms of the @xmath27 distributions for different data sets are shown in fig . [
fig : differential ] .
the complementary cumulative distributions of the data are given in table [ tab : cl ] and shown in fig .
[ fig : cumulative ] .
@llllll|rr @xmath47 & 1 & 2 & 3 & 5 & 10 & @xmath48&@xmath49 + ' '' '' interlab & @xmath50 & @xmath51 & @xmath52 & @xmath53 & @xmath54 & @xmath55 & @xmath56 + ( key ) & @xmath57 & @xmath52 & @xmath58 & @xmath59 & @xmath60 & @xmath61 & @xmath62 + nuclear & @xmath63 & @xmath64 & @xmath65 & @xmath66 & @xmath67 & @xmath68 & @xmath69 + particle & @xmath70 & @xmath64 & @xmath71 & @xmath72 & @xmath73 & @xmath74 & @xmath75 + ( stable ) & @xmath76 & @xmath77 & @xmath66 & @xmath78 & @xmath79 & @xmath80 & @xmath81 + medical & @xmath82 & @xmath83 & @xmath84 & @xmath85 & @xmath86 & @xmath87 & @xmath88 + constants & @xmath89 & @xmath90 & @xmath91 & @xmath92 & @xmath93 & @xmath94 & @xmath95 + ' '' '' normal ( gaussian ) & @xmath96 & @xmath97 & @xmath98 & @xmath99 & @xmath100 & @xmath101 & @xmath102 + student - t ( @xmath103 ) & @xmath104 & @xmath105 & @xmath106 & @xmath107 & @xmath108 & @xmath109 & @xmath110 + exponential & @xmath111 & @xmath91 & @xmath112 & @xmath78 & @xmath113 & @xmath114 & @xmath115 + student - t ( @xmath116 ) & @xmath89 & @xmath83 & @xmath117 & @xmath118 & @xmath119 & @xmath120 & @xmath121 + cauchy & @xmath122 & @xmath123 & @xmath124 & @xmath58 & @xmath125 & @xmath126 & @xmath127 + the observed probability of two measurements disagreeing by more than @xmath27 standard uncertainties for different data sets : @xmath128 .
( see also table [ tab : cl ] ) ] none of the data are close to gaussian , but all can reasonably be described by almost - cauchy student - t distributions with @xmath129 . for comparison ,
fits to these data with lvy stable distributions have nominal @xmath42 4 to 30 times worse than the fits to student - t distributions .
the number of `` @xmath0 '' ( i.e. @xmath130 ) disagreements observed is as high as @xmath53 , compared to the @xmath131 expected for a normal distribution .
the fitted values for @xmath37 and @xmath41 are shown in table [ tab : fits ] .
also shown in table [ tab : fits ] are two data subsets expected to be of higher quality , bipm interlaboratory key comparisons ( 372 quantities , 3712 measurements , 20245 pairs ) and stable particle properties ( 335 quantities , 3041 measurements , 16649 pairs ) . the key comparisons @xcite should define state - of - the - art accuracy , since they are measurements of important metrological standards carried out by national laboratories .
stable particles are often easier to study than other particles , so their properties are expected to be better determined .
both `` better '' data subsets do have narrower distributions consistent with higher quality , but they still have heavy tails .
more selected data subsets are discussed in section [ ssec : selected subsets ] . [ cols="<,^,^,^,^,^,^,^,^,^ " , ] the key metrology data subset is for electrical , radioactivity , length , mass , and other similar physical metrology standards . to see
if the most experienced national laboratories were more consistent , table [ tab : special ] also lists selected metrology data from only the six national labs that reported the most key metrology measurements . these laboratories were ptb ( physikalisch - technische bundesanstalt , germany ) , nmij ( national metrology institute of japan ) , nist ( national institutes of standards and technology , usa ) , npl ( national physical laboratory , uk ) , nrc ( national research council , canada ) , and lne ( laboratoire national de mtrologie et dessais , france ) .
similarly , key analytical chemistry data selected from the same national labs are also shown .
these are for measurements such as the amount of mercury in salmon , pcbs in sediment , or chromium in steel .
the metrology measurements by the selected national laboratories do have much lighter tails with @xmath132 , but this is not the case for their analytical measurements where @xmath133 .
new stable particle data have the lightest tail in table [ tab : fits ] , but it is not clear if this is because the newer results have better determined uncertainties or are just more correlated . the trend in particle physics is for fewer but larger experiments , and more than a third of the newer stable measurements were made by just two very similar experiments ( belle and babar ) , so the new stable data is split into two groups in table [ tab : special ] .
there is no significant difference between the belle / babar and other experiments data .
nuclear lifetimes with small and large relative uncertainties were compared .
they have similar tails , but the smaller uncertainty measurements appear to underestimate their uncertainty scales . measurements of newton s gravitation constant are notoriously variable @xcite , so a data - set without @xmath134 results was examined .
the heavy tail is reduced , albeit with large uncertainty .
the accuracy of uncertainty evaluations appears to be similar in all fields , but unsurprisingly there are noticeable differences in the relative sizes of the uncertainties . in particular , although individual physics measurements are not typically more reproducible than in medicine , they often have smaller relative uncertainty ( i.e. uncertainty / value ) as shown in fig .
[ fig : precision ] .
distribution of the relative uncertainty for data from fig .
[ fig : differential ] . ]
median ratio of the relative uncertainties ( newer / older ) for measurements in each @xmath27 pair as a function of the years between the two measurements : medical ( brown circles ) , particle ( green triangles ) , nuclear ( red squares ) , stable ( green dashed point - down triangles ) , constants ( orange diamonds ) . ]
perhaps more importantly for discovery reproducibility , uncertainty improves more rapidly in physics than in medicine , as is shown in fig .
[ fig : precisionimprovement ] .
this difference in rates of improvement reflects the difference between measurements that depend on steadily evolving technology versus those using stable methods that are limited by sample sizes and heterogeneity @xcite .
the expectation of reduced uncertainty in physics means that it is feasible to take a wait - and - see attitude towards new discoveries , since better measurements will quickly confirm or refute the new result .
measurement uncertainty in nuclear and particle physics typically improves by about a factor of 2 every 15 years .
constants data improve twice as fast , which is unsurprising since more effort is expected for more important quantities .
physicists also tend not to make new measurements unless they are expected to be more accurate than previous measurements . in the data sets reported here , the median improvement in uncertainty of nuclear measurements compared to the best previous measurement of the same quantity is @xmath135 , and the improvement factors for constants , particle , and stable measurements are @xmath136 , @xmath137 , and @xmath138 .
in contrast , medical measurements typically have greater uncertainties than the best previous measurements , with median @xmath139 .
this is an understandable consequence of different uncertainty to cost relationships in physics and medicine .
study population size is a major cost driver in medical research , so reducing the uncertainty by a factor of two can cost almost four times as much , which is rarely the case in physics .
prior expectations exist for most measurements reported here except for the interlab data .
such expectations may suppress heavy tails by discouraging publication of the anomalous results that populate the tails , since before publishing a result dramatically different from prior results or theoretical expectations , researchers are likely to make great efforts to ensure that they have not made a mistake .
journal editors , referees and other readers also ask tough questions of such results , either preventing publication or inducing further investigation . for example , initial claims @xcite of @xmath140 evidence for faster - than - light neutrinos and cosmic inflation did not survive to actual publication @xcite .
median @xmath27 value as a function of time difference between the two measurements in each @xmath27 pair : medical ( brown circles ) , particle ( green point - up triangles ) , nuclear ( red squares ) , constants ( orange diamonds ) , and stable ( green dashed point - down triangles ) . ] fig .
[ fig : correlation ] shows that physics ( particle , nuclear , constants ) measurements are more likely to agree if the difference in their publication dates is small .
such `` bandwagon effects '' @xcite are not observed in the medical data , and they are irrelevant for interlab quantities which are usually measured almost simultaneously .
these correlations imply that measurements are biased either by expectations or common methodologies .
such correlations might explain the small ( @xmath141 ) values of @xmath142 for nuclear , particle , and constants data , or it could be that researchers in these fields simply tend to overestimate the scale of their uncertainties @xcite .
removing expectation biases from the physics data would likely make their tails heavier . although interlab data are not supposed to have any expectation biases , they are subject to methodological correlations due to common measurement models , procedures , and types of instrumentation , so even their tails would likely increase if all measurements could be made truly independent .
in a famous dispute with cauchy in 1853 , eminent statistician irne - jules bienaym ridiculed the idea that any sensible instrument had cauchy uncertainties @xcite .
a century later , however , harold jeffreys noted that systematic errors may have a significant cauchy component , and that the scale of the uncertainty contributed by systematic effects depends on the size of the random errors @xcite .
the results of this study agree with earlier research that also observed student - t tails , but only looked at a handful of subatomic or astrophysics quantities up to @xmath143 @xcite .
unsurprisingly , the tails reported here are mostly heavier than those reported for repeated measurements made with the same instrument ( @xmath144 ) @xcite , which should be closer to normal since they are not independent and share most systematic effects . instead of student - t tails , exponential tails have been reported for several nuclear and particle physics data sets @xcite , but in all cases some measurements were excluded . for example , the largest of these studies @xcite looked at particle data ( 315 quantities , 53322 pairs ) using essentially the same method as this paper , but rejected the 20% of the data that gave the largest contributions to the @xmath42 for each quantity , suppressing the heaviest tails . despite this data selection , all these studies have supra - exponential heavy tails for @xmath145 , and so are qualitatively consistent with the results of this paper .
it is possible that averaging different quantities with exponential tails might produce apparent power - laws @xcite , but this would require wild variations in the accuracy of the uncertainty estimates . instead of looking directly at the shapes of the measurement consistency distributions , hedges @xcite compared particle physics and psychology results and found them to have similar compatibility , with typically almost half of the quantities in both fields having statistically significant disagreements .
thompson and ellison reported substantial amounts of `` dark uncertainty '' in chemical analysis interlaboratory comparisons @xcite .
uncertainty is `` dark '' if it does not appear as part of the known contributions to the uncertainty of individual measurements , but is inferred to exist because the dispersion of measured values is greater than expected based on the reported uncertainties .
for example , six ( 21% ) of 28 bipm key comparisons studied had ratios ( @xmath146 ) of expected to observed standard deviations less than 0.2 .
this agrees with the key analytical results in table [ tab : special ] ( which include some of the same key comparisons ) . for sample sizes matching the 28 comparisons , 20% of samples drawn from a @xmath147 student - t distribution
would be expected to have @xmath148 .
the open science collaboration ( osc ) recently replicated 100 studies in psychology @xcite , providing some of the most direct evidence yet for poor scientific reproducibility . using the osc study s supplementary information
, @xmath27 can be calculated for 87 of the reported original / replication measurement pairs , and 27 ( 31% ) disagree by more than @xmath149 , and 2 ( 2.3% ) by more than @xmath1 .
this rate of disagreements is inconsistent with selection bias acting on a normal distribution unless the @xmath150 data are excluded , but can be explained by selection biased student - t data with @xmath151 , consistent with the medical data reported in table [ tab : fits ] . when a measurement turns out to be wrong , the reasons for this failure are often unknown , or at least unpublished , so it is interesting to look at examples where the causes were later understood or can be inferred
. for medical research , heterogeneity in methods or populations is a major source of variance .
the largest inconsistency in the medical dataset is in a comparison of fever rates after acellular versus whole - cell pertussis vaccines @xcite .
the large variance can likely be explained by significant differences among the study populations and especially in how minor adverse events were defined and reported .
the biggest @xmath27 values in the particle data come from complicated multi - channel partial wave analyses of strong scattering processes , where many dozens of quantities ( particle masses , widths , helicities , ) are simultaneously determined .
significant correlations often exist between the fitted values of the parameters but are not always clearly reported , and evaluations may not always include the often large uncertainties from choices in data and parameterization .
the largest disagreement in the interlab data appears to be an obvious mistake . in a comparison of radioactivity in water @xcite ,
one lab reported an activity of @xmath152 bq / kg when the true value was about 31 . even without knowing the expected activity , the unreasonably small fractional uncertainty should probably have flagged this result .
such gross errors can produce almost - cauchy deviations .
for example , if the numerical result of a measurement is simply considered as an infinite bit string , then any `` typographical '' glitch that randomly flips any bit with equal probability will produce deviations with a @xmath153 distribution .
one can hope that the best research will not be sloppy , but not even the most careful scientists can avoid all unpleasant surprises . in 1996 a team from ptb ( the national metrological institute of germany ) reported a measurement of @xmath134 that differed by @xmath154 from the accepted value ; it took 8 years to track down the cause a plausible but erroneous assumption about their electrostatic torque transmitter unit @xcite .
a @xmath155 difference between the codata2006 and codata2010 fine structure constant values was due to a mistake in the calculation of some eighth - order terms in the theoretical value of the electron anomalous magnetic moment @xcite . a 1999 determination @xcite of avogadro s number by a team from japan s national research laboratory of metrology using the newer x - ray crystal density method was off by @xmath156 due to subtle silicon inhomogeneities @xcite . in an interlaboratory comparison measuring pcb contamination in sediments , the initial measurement by bam ( the german federal institute for materials research and testing ) disagreed by many standard uncertainties , but
this was later traced to cross - contamination in sample preparation @xcite .
several nuclear half - lives measured by the us national institute for standards and technology were known for some years to be inconsistent with other measurements ; it was finally discovered that a nist sample positioning ring had been slowly slipping over 35 years of use @xcite .
often discrepancies are never understood and are simply replaced by newer results .
for example , despite bringing in a whole new research team to go over every component and system , the reason for a discordant nist measurement of planck s constant was never found , but newer measurements by the same group were not anomalous @xcite .
heavy tails have many potential causes , including bias @xcite , overconfident uncertainty underestimates @xcite , and uncertainty in the uncertainties @xcite , but it is not immediately obvious how these would produce the observed t distributions with so few degrees of freedom . even when the uncertainty @xmath157 is evaluated from the standard deviation of multiple measurements from a normal distribution so that a student - t distribution would be expected , there are typically so many measurements that @xmath37 should be much larger than what is observed . exceptions to this are when calibration uncertainties dominate , since often only a few independent calibration points are available , or when uncertainties from systematic effects are evaluated by making a few variations to the measurements , but these can not explain most of the data .
any reasonable publication bias applied to measurements with gaussian uncertainties can not create very heavy tails , just a distorted distribution with gaussian tails to produce one false published @xmath158 result would require bias strong enough to reject millions of studies . underestimating @xmath41 does not produce a heavy tail , only a broader normal @xmath27 distribution . mixing multiple normal distributions does not naturally produce almost - cauchy distributions , except in special cases such as the ratio of two zero - mean gaussians .
the heavy tails are not caused by poor older results .
the heaviest - tailed data in fig .
[ fig : differential ] are actually the newest
93% of the interlaboratory data are less than 16 years old and eliminating older results taken prior to the year 2000 does not reduce the tails for most data as shown in table [ tab : fits ] .
intentionally making up results , i.e. fraud , could certainly produce outliers , but this is unlikely to be a significant problem here .
since most of the data were extracted from secondary meta - analyses ( e.g. review of particle properties , table of radionuclides , and cochrane systematic reviews ) , results withdrawn for misconduct prior to the time of the review would likely be excluded .
one meta - analysis in the medical dataset does include studies that were later shown to be fraudulent @xcite , but the fraudulent results actually contribute slightly less than average to the overall variance among the results for that meta - analysis .
modelling the heavy tails may help us understand the observed distributions .
one way is to assume that the measurement values are normally distributed with standard deviation @xmath159 that is unknown but which has a probability distribution @xmath160 @xcite .
the measured value @xmath33 is then expected to have a probability distribution @xmath161 this is essentially a bayesian estimate with prior @xmath162 and a normal likelihood with unknown variance .
if the uncertainties are accurately evaluated and normal with variance @xmath163 , @xmath162 will be a narrow peak at @xmath164 . assuming that @xmath162 is a broad normal distribution leads to exponential tails @xcite for large @xmath27 .
in order to generate student - t distributions , @xmath162 must be a scaled inverse chi - squared ( or gamma ) distribution in @xmath165 @xcite .
this works mathematically , but why would variations in @xmath41 for independent measurements have such a distribution ?
heavy tails can only be generated by effects that can produce a wide range of variance , so we must model how consistency testing is used by researchers to constrain such effects .
consistency is typically tested using a metric such as the calculated chi - squared statistic for the agreement of @xmath166 measurements @xmath167 @xcite @xmath168 where @xmath169 is the @xmath167 weighted mean and @xmath170 are the standard uncertainties reported by the researchers . for accurate standard uncertainties
, @xmath171 will have a chi - squared probability distribution with .
if , however , the reported uncertainties are incorrect and the true standard uncertainties are @xmath172 , then it will be @xmath173 that is chi - squared distributed .
researchers will likely search for problems if different consistency measurements have a poor @xmath174 , which typically means @xmath175 .
the larger an unknown systematic error is , the more likely it is to be detected and either corrected or included in the reported uncertainty , so published results typically have @xmath176 .
since @xmath177 is expected to have a chi - squared distribution , a natural prior for @xmath165 is indeed the scaled inverse chi - squared distribution needed to generate student - t distributions from equation [ eq : shlyakhter ] .
more mechanistically , it could be assumed that a normally distributed systematic error will be missed by @xmath178 independent measurements if their @xmath179 is less than some threshold @xmath180 .
if the distribution of all possible systematic effects is @xmath181 , then the probability distribution for the unfound errors will be @xmath182 where @xmath183 is the cumulative @xmath42 distribution .
@xmath181 is unknown , but a common bayesian scale - invariant choice is @xmath184 , with @xmath185 . using this model with the reported uncertainty @xmath41 as the lower integration bound , the curve generated from equations [ eq : shlyakhter ] and [ eq : model ] is very close to a @xmath186 student - t distribution .
the observed small values for @xmath37 mean that both @xmath178 and @xmath187 must be small .
making truly independent consistency tests is difficult , so it is not surprising that the effective number of checks ( @xmath178 ) is usually small .
this model is plausible , but why are systematic effects consistent with a @xmath184 power - law size distribution ?
scientific measurements are made by complex systems of people and procedures , hardware and software , so one would expect the distribution of scientific errors to be similar to those produced by other comparable systems .
power - law behaviour is ubiquitous in complex systems @xcite , with the cumulative distributions of observed sizes ( @xmath188 ) for many effects falling as @xmath189 , and these heavy tails exist even when the system has been designed and refined for optimal results .
a student - t distribution has cumulative tail exponent @xmath190 , and the values for @xmath37 reported here are consistent with power - law tails observed in other designed complex systems .
the frequency of software errors typically has a cumulative power - law tail corresponding to small @xmath191 @xcite , and in scientific computing these errors can lead to quantitative discrepancies orders of magnitude greater than expected @xcite .
the size distribution of electrical power grid failures has @xmath192 @xcite , and the frequency of spacecraft failures has @xmath193 @xcite .
even when designers and operators really , really want to avoid mistakes , they still occur : the severity of nuclear accidents falls off only as @xmath194 @xcite , similar to the power - laws observed for the sizes of industrial accidents @xcite and oil spills @xcite .
some complex medical interventions have power - law distributed outcomes with @xmath195 @xcite . combining the observed power - law responses of complex systems with the power - law constraints of consistency checking for systematic effects discussed in section [ ssec : modelling ]
, leads naturally to the observed consistency distributions with heavy power - law tails .
there are also several theoretical arguments that such distributions should be expected .
a systematic error or mistake is an example of a risk analysis incident , and power - law distributions are the maximal entropy solutions for such incidents when there are multiple nonlinear interdependent causes @xcite , which is often the case when things go wrong in research .
scientists want to make the best measurements possible with the limited resources they have available , so scientific research endeavours are good examples of highly structured complex systems designed to optimize outcomes in the presence of constraints .
such systems are expected to exhibit `` highly optimized tolerance '' @xcite , being very robust against designed - for uncertainties , but also hypersensitive to unanticipated effects , resulting in power - law distributed responses .
simple continuous models for highly optimized tolerant systems are consistent with the heavy tails observed in this study .
these models predict that @xmath196 @xcite , where @xmath197 is the effective dimensionality of the system , but larger values of @xmath187 arise when some of the resources are used to avoid large deviations @xcite , e.g. spending time doing consistency checks .
if one believes that mistakes can be eliminated and all systematic errors found if we just work hard enough and apply the most rigorous methodological and statistical techniques , then results from the best scientists should not have heavy tails .
such a belief , however , is not consistent with the experienced challenges of experimental science , which are usually hidden in most papers reporting scientific measurements @xcite .
as beveridge famously noted @xcite , often everyone else believes an experiment more than the experimenters themselves .
researchers always fear that there are unknown problems with their work , and traditional error analysis can not `` include what was not thought of '' @xcite .
it is not easy to make accurate _ a priori _ identifications of those measurements that are so well done that they avoid having almost - cauchy tails .
expert judgement is subject to well - known biases @xcite , and obvious criteria to identify better measurements may not work .
for example , the open science collaboration found that researchers experience or expertise did not significantly correlate with the reproducibility of their results @xcite the best predictive factor was simply the statistical significance of the original result .
the best researchers may be better at identifying problems and not making mistakes , but they also tend to choose the most difficult challenges that provide the most opportunities to go wrong . reducing heavy tails
is challenging because complex systems exhibit scale invariant behaviour such that reducing the size of failures does not significantly change the shape of their distribution . improving sensitivity makes previously unknown small systematic issues visible so they can be corrected or included in the total uncertainty .
this improvement reduces @xmath41 , but even smaller systematic effects now become significant and tails may even become heavier and @xmath37 smaller . comparing figures [ fig : differential ] and [ fig : precision ]
, it appears that data with higher average relative uncertainty tend to have heavier tails .
this relationship between relative uncertainty and measurement dispersion is reminiscent of the empirical horwitz power - law in analytical chemistry @xcite , where the relative spread of interlaboratory measurements increases as the required sensitivity gets smaller , and of taylor s law in ecology where the variance grows with sample size so that the uncertainty on the mean does not shrink as @xmath198 @xcite . in principle
, statistical errors can be made arbitrarily small by taking enough data , and @xmath37 can be made arbitrarily large by making enough independent consistency checks , but researchers have only finite time and resources so choices must be made .
taking more consistency check data limits the statistical uncertainty , since it is risky to treat data taken under different conditions as a single data set .
consistency checks are never completely independent since it is impossible for different measurements of the same quantity not to share any people , methods , apparatus , theory or biases , so researchers must decide what tests are reasonable .
the observed similar small values for @xmath37 may reflect similar spontaneous and often unconscious cost - benefit analyses made by researchers . the data showing the lightest tail reported here ( in table [ tab : special ] )
may provide some guidance and caution .
the high quality of the selected metrology standards measurements by leading national laboratories shows that heavy tails can be reduced by collaboratively taking great care to ensure consistency by sharing methodology and making regular comparisons .
there are , however , limits to what can be achieved , as illustrated by the much heavier tail of the analytical standards measured by the same leading labs .
second , consistency is easier than accuracy .
interlaboratory comparisons typically take place over relatively short periods of time , with participating institutions using the best standard methods available at that time .
biases in the standard methods may only be later discovered when new methods are introduced .
for example , work towards a redefinition of the kilogram and the associated development of new silicon atom counting technology revealed inconsistencies with earlier watt - balance measurements , and this has driven improvements in both methods @xcite . finally , selection bias that hides anomalous results is hard to eliminate . for one metrology key comparison ,
results from one quantity were not published because some laboratories reported `` incorrect results '' @xcite . reducing tails is particularly challenging for measurements where the primary goal is improved sensitivity that may lead to new scientific understanding . by definition ,
a better measurement is not an identical measurement , and every difference provides room for new systematic errors , and every improvement that reduces the uncertainty makes smaller systematic effects more significant .
frontier measurements are always likely to have heavier tails .
published scientific measurements typically have non - gaussian almost - cauchy @xmath199 error distributions , with up to @xmath200 of results in disagreement by @xmath150 .
these heavy tails occur in even the most careful modern research , and do not appear to be caused by selection bias , old inaccurate data , or sloppy measurements of uninteresting quantities . for even the best scientists working on well understood measurements using similar methodology , it appears difficult to achieve consistency better than @xmath201 , with about @xmath202 of results expected to be @xmath150 outliers , a rate a thousand times higher than for a normal distribution . these may , however , be underestimates . because of selection / confirmation biases and methodological correlations , historical consistency can only set lower bounds on heavy tails
multiple measurements may all agree but all be ( somewhat ) wrong .
the effects of unknown systematic problems are not completely unpredictable .
scientific measurement is a complex process and the observed distributions are consistent with unknown systematics following the low - exponent power - laws that are theoretically expected and experimentally observed for fluctuations and failures in almost all complex systems .
researchers do determine the scale of their uncertainties with fair accuracy , with the scale of medical uncertainties ( @xmath203 ) slightly more consistent with the expected value ( @xmath204 ) than in physics ( @xmath205 ) .
medical and physics research have comparable reproducibility in terms of how well different studies agree within their uncertainties , consistent with a previous comparison of particle physics with social sciences @xcite .
medical research may have slightly lighter tails , while physics results typically have better relative uncertainty and greater statistical significance .
understanding that error distributions are often almost - cauchy should encourage use of t - based @xcite , median @xcite , and other robust statistical methods @xcite , and supports choosing student - t @xcite or cauchy @xcite priors in bayesian analysis .
outlier - tolerant methods are already common in modern meta - analysis , so there should be little effect on accepted values of quantities with multiple published measurements , but this better understanding of the uncertainty may help improve methods and encourage consistency .
false discoveries are more likely if researchers apply normal conventions to almost - cauchy data . although much abused , the historically common use of @xmath206 as a discovery criterion suggests that many scientists would like to be wrong less than 5% of the time . if so , the results reported here support the nominal 5-sigma discovery rule in particle physics , and may help discussion of more rigorous significance criteria in other fields @xcite . this study should help researchers better understand the uncertainties in their measurements , and may help decision makers and the public better interpret the implications of scientific research @xcite . if nothing else , it should also remind everyone to never use normal / gaussian statistics when discussing the likelihood of extreme results .
i thank the students of the university of toronto advanced undergraduate physics lab for inspiring consideration of realistic experimental expectations , and the university of auckland physics department for their hospitality during very early stages of this work .
i am grateful to d. pitman for patient and extensive feedback , to r. bailey for his constructive criticism , to r. cousins , d. harrison , j. rosenthal , and p. sinervo for useful suggestions and discussion , and to m. cox for many helpful comments on the manuscript .
the sources for all data analysed are listed in the associated ancillary file : uncertaintydatadescription.xls .
rosenfeld ah .
1968 are there any far - out mesons or baryons ? in _
meson spectroscopy _ ( eds baltay c , rosenfeld ah ) .
new york , usa : w. a. benjamin .
. 455483 .
https://archive.org/details/mesonspectroscopy [ ] nakagawa s , cuthill ic .
2007 effect size , confidence interval and statistical significance : a practical guide for biologists .
rev . _ * 82 * , 591605 . http://dx.doi.org/10.1111/j.1469-185x.2007.00027.x [ ] sinervo pk .
2003 definition and treatment of systematic uncertainties in high energy physics and astrophysics . in _
phystat 2003 , stanford , usa _
( eds lyons l , mount rp , reitmeyer r ) .
http://stanford.io/2fdpkta [ ] bravin e , brun g , dehning b , drees a , galbraith p , geitz m , henrichsen k , koratzinos m , mugnai g , tonutti m. 1998 the influence of train leakage currents on the lep dipole field . _ nucl .
instrum . meth . a _ * 417 * , 915 .
http://dx.doi.org/10.1016/s0168-9002(98)00020-5 [ ] attivissimo f , cataldo a , fabbiano l , giaquinto n. 2011 systematic errors and measurement uncertainty : an experimental approach .
_ measurement _ * 44 * , 17811789 .
http://dx.doi.org/10.1016/j.measurement.2011.07.011 [ ]
willink r. 2004 approximating the difference of two t - variables for all degrees of freedom using truncated variables .
n. z. j. stat . _
* 46 * , 495504 .
[ ] faller je .
2014 precision measurement , scientific personalities and error budgets : the sine quibus non for big g determinations .
r. soc . a _ * 372 * , 20140023 .
http://dx.doi.org/10.1098/rsta.2014.0023 [ ]
bienaym m. 1853 considrations a lappui de la dcourverte de laplace sur la loi de probabilit dans la mthode des moindres carrs . _ c. r. acad .
sci . _ * xxxvii * , 309324 .
https://books.google.com/books?id=qhjfaaaacaaj&pg=pa322 [ ]
zhang l , prietsch so , axelsson i , halperin sa .
2012 acellular vaccines for preventing whooping cough in children . _
cochrane database of systematic reviews _
http://dx.doi.org/10.1002/14651858.cd001478.pub5 [ ] .
2010 almera proficiency test determination of naturally occurring radionuclides in phosphogypsum and water ( iaea - cu-2008 - 04 ) .
austria : international atomic energy agency iaea / aq/15 . http://www-pub.iaea.org/mtcd/publications/pdf/iaea-aq-15_web.pdf [ ] fujii k , tanaka m , nezu y , nakayama k , fujimoto h , de bievre p , valkiers s. 1999 determination of the avogadro constant by accurate measurement of the molar volume of a silicon crystal .
_ metrologia _ * 36 * , 455464 . http://dx.doi.org/10.1088/0026-1394/36/5/7 [ ] carlisle j , pace n
, cracknell j , mller a , pedersen t , zacharias m. 2013 ( 5 ) what should the cochrane collaboration do about research that is , or might be , fraudulent ? _ cochrane database of systematic reviews_. http://dx.doi.org/10.1002/14651858.ed000060 [ ] dose v , von der linden w. 1999 outlier tolerant parameter estimation . in _
maximum entropy and bayesian methods : garching , germany 1998 _ vol .
105 of _ fundamental theories of physics_. netherlands : springer .
. 4756 .
http://dx.doi.org/10.1007/978-94-011-4710-1_4 [ ]
dobson i , carreras ba , lynch ve , newman de .
2007 complex systems analysis of series of blackouts : cascading failure , critical points , and self - organization .
_ chaos _ * 17 * , 026103 .
http://dx.doi.org/10.1063/1.2737822 [ ] karimova lm , kruglun oa , makarenko ng , romanova nv .
2011 power law distribution in statistics of failures in operation of spacecraft onboard equipment .
res . _ * 49 * , 458463 .
http://dx.doi.org/10.1134/s0010952511040058 [ ] sornette d , maillart t , kroeger w. 2013 exploring the limits of safety analysis in complex technological systems .
. j. disaster risk reduct . _ * 6 * , 5966 . | judging the significance and reproducibility of quantitative research requires a good understanding of relevant uncertainties , but it is often unclear how well these have been evaluated and what they imply .
reported scientific uncertainties were studied by analysing 41000 measurements of 3200 quantities from medicine , nuclear and particle physics , and interlaboratory comparisons ranging from chemistry to toxicology .
outliers are common , with @xmath0 disagreements up to five orders of magnitude more frequent than naively expected
. uncertainty - normalized differences between multiple measurements of the same quantity are consistent with heavy - tailed distributions that are often almost cauchy , far from a gaussian normal bell curve .
medical research uncertainties are generally as well evaluated as those in physics , but physics uncertainty improves more rapidly , making feasible simple significance criteria such as the @xmath0 discovery convention in particle physics .
contributions to measurement uncertainty from mistakes and unknown problems are not completely unpredictable .
such errors appear to have power - law distributions consistent with how designed complex systems fail , and how unknown systematic errors are constrained by researchers .
this better understanding may help improve analysis and meta - analysis of data , and help scientists and the public have more realistic expectations of what scientific results imply . |
the idea that small observable violations of lorentz symmetry could provide experimental access to planck - scale effects continues to draw attention across several subfields of physics . in the three years since the previous meeting in this series , considerable progress has been made on both experimental and theoretical fronts .
this contribution to the cpt13 proceedings contains a brief introduction , followed by comments on two topics of recent interest : nonminimal fermion couplings and riemann - finsler geometry .
a satisfactory theoretical description of lorentz violation must incorporate coordinate independence , realism , and generality .
a powerful approach uses effective field theory , starting with general relativity coupled to the standard model and adding to the lagrange density all observer - invariant terms that contain lorentz - violating operators combined with controlling coefficients .
this yields the comprehensive realistic effective field theory for lorentz violation called the standard - model extension ( sme ) @xcite .
the sme also describes general cpt violation , which in the context of realistic effective field theory is accompanied by lorentz violation @xcite .
the full sme contains operators of arbitrary mass dimension @xmath0 , while the minimal sme restricts attention to operators of renormalizable dimension @xmath1 .
observable signals of lorentz violation are governed by the sme coefficients .
experiments typically search for particle interactions with background coefficient values , which can produce effects dependent on the particle velocity , spin , flavor , and couplings .
many investigations of this type have been performed @xcite , achieving impressive sensitivities that in some cases exceed expectations for suppressed planck - scale effects . if the sme coefficients are produced by spontaneous lorentz violation @xcite , as is necessary when gravity is based on riemann geometry @xcite , then they are dynamical quantities that must incorporate massless nambu - goldstone modes @xcite .
these modes have numerous interpretations , including serving as an alternative origin for the photon in einstein - maxwell theory @xcite and the graviton in general relativity @xcite , or representing new spin - dependent @xcite or spin - independent @xcite forces , among other possibilities .
massive modes can also appear @xcite .
in the nonminimal sector of the sme , the number of lorentz - violating operators grows rapidly with the mass dimension @xmath0 . systematically cataloguing and characterizing the possibilities
is therefore indispensible in the search for lorentz violation . in the cpt10 proceedings
, i outlined some features appearing in the treatment of quadratic operators of arbitrary @xmath0 in the photon lagrange density @xcite . in the intervening three - year period ,
investigations of the quadratic fermion sector for arbitrary @xmath0 have also been performed .
the lagrange density for propagation and mixing of any number of fermions has been developed and applied to describe general lorentz violation in the neutrino sector @xcite .
more recently , quadratic operators of arbitrary @xmath0 have been studied for a massive dirac fermion @xcite .
many nonminimal operators generate effects that are in principle observable , and each such operator generates a distinct experimental signal . for quadratic operators , which characterize particle propagation and phase - space features of particle decays , the lorentz - violating effects can include direction dependence ( anisotropy ) , wave - packet deformation ( dispersion ) , and mode splitting ( birefringence ) . in the neutrino sector , for example , some operators control flavor - dependent effects in neutrino and antineutrino mixing , producing novel energy and direction dependences that involve both dirac- and majorana - type couplings .
others govern species - independent effects , which can differ for neutrinos and antineutrinos and can produce propagation times varying with energy and direction , in some cases exceeding that of light .
a few operators produce ` countershaded ' effects@xcite that can not be detected via oscillations or propagation but change interaction properties in processes such as beta decay @xcite .
analogous effects appear in the description of a massive dirac fermion in the presence of lorentz violation , for which the exact dispersion relation for arbitrary @xmath0 is known in closed and compact form @xcite .
for example , the fermion group velocity is anisotropic and dispersive , while the fermion spin exhibits a larmor - like precession caused by birefringent operators . using field redefinitions to investigate observability reveals that many operators of dimension @xmath0 produce no effects or are physically indistinguishable from others of dimensions @xmath0 or @xmath2 .
nonetheless , the number of observable coefficients grows as the cube of @xmath0 . to date , almost all the nonminimal coefficient space for fermions is experimentally untouched .
this offers an open arena for further exploration with a significant potential for discovery .
the surprising ` no - go ' result that the conventional riemann geometry of general relativity and its extension to riemann - cartan geometry are both incompatible with explicit lorentz violation raises the questions of whether an alternative geometry is involved and , if so , whether a corresponding gravitational theory exists .
the obstruction to explicit lorentz violation , which disappears for the spontaneous case , can be traced to the generic incompatibility of the bianchi identities with the external prescription of coefficients for lorentz violation .
it is therefore reasonable to conjecture that a natural geometrical setting would include metric distances depending locally on the coefficients in addition to the riemann metric .
@xcite support for this conjecture has recently emerged with the discovery that a fermion experiencing explicit lorentz violation tracks a geodesic in a pseudo - riemann - finsler geometry rather than a conventional geodesic in pseudo - riemann spacetime @xcite .
riemann - finsler geometry is a well - established mathematical field with numerous physical applications ( see , e.g. , ref . ) , such as the famous zermelo navigation problem of obtaining the minimum - time path for a ship in the presence of ocean currents .
a large class of riemann - finsler geometries is determined by the geodesic motion of particles in the sme @xcite . among these
are the canonical randers geometry , which is related to the 1-form sme coefficient @xmath3 , and numerous novel geometries of simplicity comparable to the randers case .
one example of the latter is @xmath4 space , a calculable riemann - finsler geometry that also is based on a 1-form and has finsler structure complementary to that of randers space .
physically , this geometry underlies the geodesic motion of a fermion in curved spacetime in the presence of chiral cpt - odd lorentz violation .
all the sme - inspired geometries exhibit mathematically interesting features connected to physical properties .
for instance , when the sme coefficients are covariantly constant , the trajectories are conventional riemann geodesics and special riemann - finsler geometries known as berwald spaces result .
many open challenges remain in this area , ranging from more technical questions such as resolving singularities or classifying geometries to physical issues such as generalizing zermelo navigation or uncovering implications for the sme .
the prospects appear promising for further insights to emerge from this geometrical approach to lorentz violation .
this work was supported in part by u.s . d.o.e .
grant de - fg02 - 13er42002 and by the indiana university center for spacetime symmetries .
kosteleck and s. samuel , phys .
d * 39 * , 683 ( 1989 ) ; phys .
lett . * 63 * , 224 ( 1989 ) ; phys .
d * 40 * , 1886 ( 1989 ) ; v.a .
kosteleck and r. potting , nucl .
b * 359 * , 545 ( 1991 ) ; phys .
d * 51 * , 3923 ( 1995 ) ; v.a .
kosteleck and r. lehnert , phys .
d * 63 * , 065008 ( 2001 ) .
d. colladay and p. mcdonald , phys .
d * 85 * , 044042 ( 2012 ) ; v.a .
kostelecky , n. russell , and r. tso , phys .
b * 716 * , 470 ( 2012 ) ; v.a .
kosteleck and n. russell , phys .
b * 693 * , 2010 ( 2010 ) . | this contribution to the cpt13 meeting briefly introduces lorentz and cpt violation and outlines two recent developments in the field . |
the genetic information in eukaryotic cells ( including cells of animals and plants ) is accessed through dna unwinding on two different length scales . on the larger scale
the dna double helix has to unwind from proteins , on the smaller scale the two strands of the double helix need to be separated . in the first case a semiflexible polymer ( dna double helix ) is wound almost two turns around a protein cylinder forming the so - called nucleosome @xcite , in the second case two flexible polymers ( chains of nucleotides ) are twisted around each other leading to the much stiffer double helix .
the unwinding of the dna from the nucleosome or the separation of the dna double helix is achieved inside a cell in various ways , often involving molecular motors ( chromatin remodellers , polymerases ... ) that usually give access through a local opening of the structures . inside a test tube unwinding
can be induced , typically on a global scale , through a change in temperature ( dna melting / helix - coil transition @xcite ) , salt concentration ( salt - induced dna release @xcite ) or through application of an external force ( nucleosome unwrapping @xcite , dna unzipping @xcite ) .
local unwinding can also occur spontaneously leading to the breathing of nucleosomes and of dna ( usually called site exposure in nucleosomes @xcite and denaturation bubbles in dna @xcite ) . finally , during transcription , where the elongating rna polymerase produces a rna transcript , one faces again the situation of a flexible chain , the transcript , being initially wound around the much stiffer dna double helix @xcite . to gain insights into the unwinding process it is convenient to start from simple setups .
in this work we study the unwinding process of an idealized flexible polymer model that is initially wound around a stiff rod , in a configuration of minimal entropy , which resembles some of the features of the dna or rna unwinding .
this model was introduced and studied in ref .
@xcite . here
we extend the results of that analysis focusing in particular on metric properties .
we present a scaling argument which fully captures the early stages of the dynamics that is characterized by power - law scaling . despite the simplicity of the model
, there is a complex dynamical behavior .
it is the entropy gain that drives the polymer from the initial configuration toward the full random coil configuration .
a sketch of this process is given in figure [ fig:1 ] , which shows different snapshots of the polymer configurations in the course of a simulation .
the polymer is initially fully wound in a helix around the rigid rod .
one end of the polymer is tethered to the rod whereas the other end is free , hence the relaxation proceeds from the free end .
this process shares some similarities with the simpler problem of the relaxation from one end of a completely stretched polymer @xcite .
consider a polymer tethered at one end and fully stretched by a strong flow .
when the flow is turned off the polymer relaxes back to its coiled equilibrium conformation . as one end is tethered the relaxation starts from a free end , from where the coil grows .
the unwinding has a similar relaxation from the free end which occurs through a rotational motion instead of a translational recoiling .
however , the phenomenology in the case of unwinding is much richer , as we will show .
this paper is organized as follows . in the section models and simulations , we review the models and the type of monte carlo simulations used . in the section results and discussion ,
we focus first on the early unwinding dynamics which is characterized by power - laws scaling in time .
it is shown that force balance equations and scaling arguments yield exponents in very good agreement with simulations .
second , we discuss the late stage of the relaxation process . here the theory of ref .
@xcite predicts a power - law scaling for the longest relaxation time with logarithmic corrections .
the analysis is extended to other lattice and off - lattice models and the results confirm the validity of the theory . [ cols="^ " , ]
figure [ fig:1 ] shows different snapshots of the polymer configurations during the unwinding from an infinitely long rod . initially the polymer is fully wound and in the course of time it unwraps . the polymer has @xmath0 monomers labeled with indices @xmath1 .
one end ( @xmath2 ) is fixed to the rod , while the other end ( @xmath3 ) is free .
we have studied different cases to check the robustness and universality of the numerical results . for the ideal chain ( the case without excluded volume )
we modeled the polymer in three different ways . in a first model we considered a random walk ( rw ) with unit steps on a face - centered cubic ( fcc ) lattice . in the initial helical configuration ,
the polymer performs one turn in six steps along the rod .
an update of the configuration consists of @xmath0 attempts of so - called corner flips , where the randomly selected monomer is moved to a neighboring lattice site if the distances to the neighboring monomers are conserved ( this is a lattice realization of rouse dynamics @xcite ) . the new configuration is accepted if the monomer does not overlap with the rod .
we also considered a random walk on a square lattice ( i.e. , a bidimensional lattice ) . in this case
the starting configuration consists of a chain with a rescaled segment length wound around the origin . in the initial configuration on the square lattice one turn around the origin
is performed by @xmath4 monomers .
the corner flip method is used here as well . finally , we used a freely jointed chain ( fjc ) to model three - dimensional off - lattice polymers .
neighboring monomers have a fixed distance @xmath5 from each other and the rod has a diameter @xmath6 ensuring that monomers can not accidentally pass from one side to the other in the course of unwinding .
the initial helical configuration is such that one turn is performed with @xmath7 monomers .
also in this case a time step of the dynamics consists in @xmath0 attempts of moves for randomly chosen monomers . when a monomer is selected , a new configuration is constructed via a rotation of that monomer around the axis defined by the two neighboring monomers by an angle randomly selected from @xmath8 $ ] .
the free end is updated with a new random position preserving the bond length with the second monomer .
the new configuration is accepted if it does not overlap with the rod .
for the case of a polymer with excluded volume we use only the model of a self - avoiding walk ( saw ) on the fcc lattice .
the procedure is the same as for the ideal chain on the fcc lattice , with the added constraint of excluded volume between the monomers : one rejects moves violating it .
we consider first the radial " distance @xmath9 which is defined as the average distance of the free end monomer from the rod . this quantity is shown in figure [ fig:2 ] ( rw s ) and [ fig:3 ] ( saw ) .
starting from a minimal value for the fully wrapped configuration at @xmath10 the growth of @xmath9 follows a power law . in the ideal chain cases we find a first , short - time , regime : @xmath11 this scaling with time of a spatial length scale is remarkably slow ( @xmath12 ) compared to the usual relaxation time scales encountered in polymer physics .
the exponent is robust and is found in the minimalistic 2d polymer on a square lattice ( figure [ fig:2](a ) ) , the rw on a fcc lattice ( figure [ fig:2](b ) ) and the freely jointed chain off - lattice ( figure [ fig:2](c ) ) .
we performed also simulations for a 3d excluded volume chain on an fcc lattice , for which we find a power law with a slightly larger exponent : @xmath13 as it can be seen in figure [ fig:3 ] . of the free end from the rod versus time for a excluded volume chain on an fcc lattice .
the short time regime follows the power law @xmath14 .
the inset shows the scaling of @xmath15 with the polymer length .
averages are made at least over @xmath16 configurations ( for the largest size).,width=264 ] to understand the origin of this exponent we model the unwinding starting from a two phase picture .
we assume that during the early stage of the dynamics the polymer starting from the fixed end has @xmath17 monomers tightly wrapped around the rod , which are frozen as in their initial @xmath10 configuration , while @xmath18 monomers are loose ( we indicate these as phase 1 and phase 2 , respectively , see figure [ fig : config](a ) ) .
we assume that the loose monomers form a homogeneous helix with a constant pace , but which is loosely wrapped around the rod .
obviously , if the loose helix would extend until the free end of the polymer , the radial distance @xmath9 would not grow in time .
the two phase model is the starting point of our analysis and we will focus on the configuration of the polymer close to the free end later .
we assume that the dynamics is governed by the following equation : @xmath19 which is a balance between frictional and `` entropic '' forces during the growth of the helical domain . here
@xmath20 is the total free energy of the configuration , with @xmath21 and @xmath22 the free energies per monomer of the two phases ( with @xmath23 as the loose helix has a higher entropy than the tight helix ) .
@xmath24 is the friction coefficient .
the total winding angle of the last monomer is equal to @xmath25 times the number of times the polymer is wrapped around the rod .
for the two helices model such a quantity is then given by : @xmath26 where @xmath27 and @xmath28 are the densities of winding for the two phases ( with @xmath29 , as phase 2 is more loosely wrapped compared to phase 1 ) .
a decrease in @xmath17 leads to a decrease in the total winding angle and the whole loose helix rotates in a corkscrew motion .
the friction coefficient is then proportional to the length of the rotating domain @xmath30 . using this input and the form of @xmath31 we get from the integration of eq .
( [ eq : domain1 ] ) : @xmath32 the angular velocity is given by eq .
( [ thetan ] ) : @xmath33 which decreases in time as there is an increasing friction when the loose helix grows .
the assumption that phase 2 is a helix of constant pace and radius is an approximation .
however , as we will show , the numerical data are in good agreement with a square root growth ( eq .
( [ growth1 ] ) ) , which is a consequence of that assumption .
note that the assumption can be relaxed , allowing phase 2 to have fluctuations ; the only essential requirement is that the friction @xmath24 should scale linearly with the length of the domain .
our approximation is similar in spirit to the monoblock approximation where an inhomogeneously stretched polymer is modeled by a homogeneously stretched one , an approximation that does not change the scaling of the large scale geometry of the deformed chain @xcite . at the polymer end .
( c - d ) view of a polymer anchored to the rod ( which is perpendicular to the plane ) and rotating with angular velocity @xmath34 .
if @xmath34 is small the polymer rotates while maintaining its equilibrium shape ( case ( c ) ) . for high @xmath34 the polymer
gets partially wrapped around the rod , while a part of length @xmath35 maintains its equilibrium shape .
equation ( [ length ] ) gives an estimate of the length @xmath35.,width=264 ] let us consider now the growth of a coiled domain at the end of the chain ( the stretch ab of length @xmath35 shown in figure [ fig : config](b ) ) .
this domain grows from a polymer rotating with angular velocity @xmath34 .
let us assume that the friction originating from the coiled part is negligible compared to that of the loose helix , so the calculations leading to eqs .
( [ growth1 ] ) and ( [ omega ] ) remain valid . to understand the coil growth we consider a polymer attached to a rod and rotating with angular velocity @xmath34 .
if the polymer is sufficiently short and @xmath34 small , its equilibrium conformation is not perturbed by the rotation and in particular it will have no winding ( figure [ fig : config](c ) ) .
if the polymer length exceeds a given threshold value , then part of the polymer close to the attachment point gets wound while the final part rotates maintaining its equilibrium shape ( length of the part ab in figure [ fig : config](d ) ) .
we estimate now the length of the end coil for a rotating polymer . to understand the calculation it is useful to consider the analogous problem of a polymer pulled by one end by a constant force @xcite .
the polymer maintains its equilibrium conformation if the applied force , @xmath36 , or the polymer length , @xmath35 , do not exceed the values fixed by the equation : @xmath37 where @xmath38 is the flory radius , @xmath39 the boltzmann s constant and @xmath40 the temperature . for a polymer rotating with an angular velocity @xmath34 the force which distorts its shape
is due to friction .
the expression analogous to eq .
( [ frf ] ) is then given by @xmath41 using @xmath42 and @xmath43 for a rouse polymer we obtain the following relation for the length of the coiled unwound end of the rotating polymer : @xmath44 using the law ( [ omega ] ) for the angular velocity we finally obtain for the growth of @xmath35 : @xmath45 and , from the equilibrium relation @xmath46 we find that the squared distance from the rod grows as : @xmath47 with @xmath48 .
note that @xmath49 for a gaussian polymer ( @xmath50 ) and @xmath51 for a self - avoiding polymer ( @xmath52 ) which is in excellent agreement with the numerical results ( figures [ fig:2 ] and [ fig:3 ] ) .
we expect that the above description remains valid until the loose helix has grown to reach the first monomer .
this corresponds to @xmath53 , i.e. when the tight helix has disappeared . according to eq .
( [ growth1 ] ) this happens at a characteristic time @xmath15 scaling as @xmath54 .
we estimated @xmath15 for polymers of different lengths from the simulation data of the radial distance of figures [ fig:2 ] and [ fig:3 ] .
this is the time at which the growth law starts deviating from eq .
( [ revst ] ) .
the vertical arrows in figure [ fig:2 ] mark the estimated @xmath15 for the polymer of shortest length .
the insets of figures [ fig:2](a ) and ( b ) show plots of @xmath15 vs. @xmath0 as obtained from the data of the main plots .
there is an excellent agreement with the predicted scaling @xmath54 .
an equally good agreement was found for all the other cases studied .
note that the scaling @xmath54 does not depend on the presence of self - avoidance , as demonstrated by the data in the inset of figure [ fig:3 ] .
we consider next the distance @xmath55 between the free end monomer at position @xmath0 and a monomer @xmath56 projected onto a plane perpendicular to the rod . in two dimensional models
this is equal to the total distance @xmath57 where @xmath58 is the position of monomer @xmath59 . in three dimensions , for a rod parallel to the @xmath60-axis we have : @xmath61 where @xmath62 and @xmath63 are the coordinates of the monomer @xmath59 . in all cases analyzed ( see figure [ fig:4 ] )
we find for small @xmath64 a scaling @xmath65 which demonstrates that the end part of the polymer is an equilibrated coil .
the plots show that @xmath66 saturates at a constant value for sufficiently large value of @xmath64 .
this point identifies the end of the coil and the beginning of the loose helical region . from scaling arguments
we expect : @xmath67 where for small values of @xmath68 the function @xmath69 converges to a constant . for large @xmath70
we expect @xmath71 as @xmath66 is independent of @xmath64 , as the projected distance from the end monomer to the monomers in the helical domain can not increase .
the inset of figure [ fig:4 ] shows a scaling plot of @xmath72 vs. @xmath73 in full agreement with the scaling ansatz ( [ scaling_ansatz ] ) .
this result supports the prediction that the coil growths according to eq .
( [ evol_length ] ) .
note that the scaling function has a maximum at the crossover between the two scaling regimes .
the maximum is not very pronounced but indicates that the polymer , compared to an equilibrated coil , is slightly more stretched in the vicinity of the rod .
the theory developed in the previous section can be tested also on the dynamics of the total winding angle .
taking into account the presence of the coil of length @xmath35 at the polymer end we need to modify eq .
( [ thetan ] ) with : @xmath74 as the coil , which is of length @xmath35 , does not contribute to the winding .
we define with @xmath75 the initial total winding angle .
combining eqs .
( [ growth1 ] ) , ( [ evol_length ] ) and ( [ thetan2 ] ) , we then get : @xmath76 with @xmath77 and @xmath78 being some positive constants .
the prediction is that @xmath79 scales as @xmath80 , with a slowly decaying correction term originating from the equilibrated end coil .
the correction is predicted to scale as @xmath81 for a rw and @xmath82 for a saw . in order to test the validity of eq .
( [ corr_scaling ] ) we plot in figure [ short_theta ] the quantity @xmath83 as a function of @xmath84 in a log - log scale for three ideal - chain models in two and three dimensions , and the same for saw s in figure [ short_theta_saw ] . in all cases
the data approach for sufficiently long times the expected @xmath80 law . to investigate the nature of the corrections to the leading scaling behavior we plot in the insets of figure [ short_theta ] the quantity @xmath85 vs. @xmath86 .
the data for short times and for different polymer lengths follow a straight line in good agreement with the prediction of eq .
( [ corr_scaling ] ) .
the slope of the lines are positive and imply @xmath87 , as expected .
some stronger nonmonotonic behavior is observed in the 3d off - lattice model which does not have a counterpart in the other cases studied .
the behavior of the winding angle was also investigated in a previous publication @xcite and estimated to scale in the early time dynamics as @xmath88 where @xmath89 .
this seemed to match the short time dynamics rather well , although a closer inspection of the data shows that eq .
( [ corr_scaling ] ) fits the data better .
the analysis of the local winding , which follows , gives further support of a @xmath90 growth of the unwound domain .
further insight of the polymer dynamics can be obtained from the analysis of the local winding angle @xmath91 , which is the winding angle of the @xmath92-th momomer . as the winding angle is counted from the monomer attached to the rod ( @xmath2 ) one has @xmath93 , whereas the total winding angle defined above
is @xmath94 .
figure [ fig : local_winding ] shows the time evolution of @xmath91 vs. @xmath92 for different times and for a polymer of length @xmath95 for a planar rw . at @xmath10
the configuration is fully wrapped around the rod which corresponds to a linear increase @xmath96 . as the time evolves @xmath91 decreases in a more pronounced manner starting from the free end of the polymer ; at short times there is a domain of length @xmath17 which is still fully wrapped as at time @xmath10 , followed by a loose part of length @xmath18 .
the inset of figure [ fig : local_winding ] shows a plot of @xmath18 vs. @xmath84 .
the data are in very good agreement with the square root growth predicted by eq .
( [ growth1 ] ) . differently from the data for the total winding angle of eq .
( [ corr_scaling ] ) , in this case there are no corrections to scaling expected . vs. monomer index @xmath92 for the 2d rw of size @xmath95 .
the dashed tilted line corresponds to the fully wrapped conformation @xmath97 .
the data are obtained for increasing times @xmath98 , @xmath99 , @xmath100 , @xmath101 , @xmath102 and @xmath103 .
the vertical dashed - dotted line denotes the boundary between the tight helix of length @xmath17 and the loose part of length @xmath18 at the time @xmath104 .
inset : the growth of the loose domain follows the square root behavior predicted by eq .
( [ growth1]).,scaledwidth=45.0% ] in the previous section we have fully characterized the early time relaxation dynamics .
we found that the first regime in which the typical polymer configuration looks like in figure [ fig : config](b ) ends at a time scaling as @xmath54 . at very long times
the dynamics was studied in ref .
@xcite using a force balance equation for the total winding angle @xmath105 .
this equation reads : @xmath106 where @xmath107 is the free energy of a polymer in equilibrium with a total winding angle @xmath105 and where @xmath108 is the friction coefficient .
the free energy is a function of a scaling variable @xmath109 , where @xmath110 is an exponent governing the fluctuations of the winding angle at equilibrium .
for rw s it is known rigorously @xcite that @xmath111 while numerical simulations of 3d saw s yield @xmath112 @xcite . for small winding angles
the free energy is quadratic in the scaling variable , so that the relaxation to equilibrium becomes exponential @xcite : @xmath113 where @xmath114 is the longest relaxation time : @xmath115 since the initial configuration is fully wound with @xmath116 , the total unwinding time is given by @xmath117 @xcite .
the data are best analyzed using the definition of a running exponent , which probes the local slope of the data in a log - log plot . from eq .
( [ taup ] ) we get : @xmath118 where we have included a scale term @xmath5 , in order to account for further corrections to scaling .
figure [ fig : eff_exp ] shows a plot of the numerical value of @xmath119 in two and three dimensions obtained from simulations of rw s .
these data extend those of ref .
@xcite by including the three dimensional case for both fcc lattice ( b ) and off - lattice ( c ) models .
the data are compared with eq .
( [ eff_exp ] ) where there is only a single adjustable parameter @xmath5 used in the fit .
the agreement is very good confirming the validity of the analytical approach of force balancing in eq .
( [ late_stage_eq ] ) , which describes the process using the total winding angle as a single reaction coordinate .
in this paper we have investigated the problem of the unwinding dynamics of a flexible polymer from a rigid rod .
scaling arguments and force balance equations allowed us to fully characterize the early stages of the unwinding and the late stages of the relaxation dynamics .
these arguments are supported by extensive numerical simulations in two and three dimensions , with and without self - avoidance .
the early dynamics can be understood by a three phase picture , where the polymer configuration starting from the fixed end can be described by a tight helix , a looser helix and a free random coil .
the latter two phases grow in time following two different dynamical laws as predicted by eqs .
( [ growth1 ] ) and ( [ evol_length ] ) .
the analysis of various quantities from numerical simulations as metric distances or winding angles are all consistent with the analytical theory .
interestingly , the first growth law ( [ growth1 ] ) does not contain the exponent @xmath120 and hence is superuniversal , being the same for random and self - avoiding walks . in the late stage dynamics
we have extended the results of ref .
@xcite to different models and confirmed the scaling form of the longest relaxation time which involves logarithmic corrections .
the emerging picture is that of a relatively quick loosening of the polymer , which remains very close to the rod in the early stages of the dynamics .
this is followed by an intermediate regime where the distance from the rod grows in a faster way and leads to the final relaxation . differently from the early and late time behaviors
the intermediate regime does not appear to display a clear cut scaling .
this can be seen , for instance , in the behavior of @xmath121 depicted in figure [ fig:2 ] . during early dynamics , until a characteristic time @xmath15 , @xmath121 follows a power - law scaling . in the late time
dynamics @xmath121 reaches a plateau .
the intermediate time regime links the two regimes showing no clear evidence of a power law behavior .
a typical snapshot of the intermediate regime is shown in figure [ fig:1 ] ( third configuration from the top ) .
its characterization remains a challenge for future work .
the two phase model description of polymer dynamics has recently gained some popularity : as examples we mention here the case of the translocation of a polymer through a nanopore @xcite , the pulling of a polymer by a constant force from one end @xcite and the folding of a dna hairpin @xcite . in these problems
the polymer is subject to some local forcing and set into motion through the propagation of tension along its backbone . as the tension does not propagate instantaneously , the polymer is not set into motion at once . to describe the motion it is usually assumed that one can divide the polymer into different phases , which leads to some analytical predictions of the exponents governing the dynamics @xcite . in this paper
a similar approach was adopted to study a complementary case , namely that of the relaxation dynamics of an initially stretched ( helically wrapped ) polymer .
the excellent agreement between simulations and model results shown in this paper , corroborates the validity of a two phase model approach in the description of nonequilibrium polymer dynamics .
we thank g.t .
barkema and a. revyakin for stimulating discussions .
this work is part of the research program of the foundation for fundamental research on matter ( fom ) , which is financially supported by the netherlands organisation for scientific research ( nwo ) .
c.w . acknowledges the support by the laboratory of excellence initiative ( labex ) numev , od by the scientific council of the university of montpellier 2 and hpc@lr for offering computational resources .
brower - toland b.d . ,
smith c.l .
, yeh r.c . , lis j.t . , peterson c.l . @xmath122
( 2002 ) mechanical disruption of individual nucleosomes reveals a reversible multistage release of dna _ proc .
usa_*99 * , 1960 - 1965 .
belotserkovskii b.p .
( 2014 ) relationships between the winding angle , the characteristic radius , and the torque for a long polymer chain wound around a cylinder : implications for rna winding around dna during transcription _ phys .
* 89 * , 022709 | we study the rotational dynamics of a flexible polymer initially wrapped around a rigid rod and unwinding from it .
this dynamics is of interest in several problems in biology and constitutes a fundamental instance of polymer relaxation from a state of minimal entropy .
we investigate the dynamics of several quantities such as the total and local winding angles and metric quantities .
the results of simulations performed in two and three dimensions , with and without self - avoidance , are explained by a theory based on scaling arguments and on a balance between frictional and entropic forces .
the early stage of the dynamics is particularly rich , being characterized by three coexisting phases . |
the _ mass inflation _ phenomenon , resulting in exponential growth of the perturbed black hole mass function at the cauchy horizon , was considered as a fatal instability of the interior kerr newman black hole solution with respect to the small perturbations ( see , e.g. , @xcite and an example of the bounded mass inflation @xcite ) .
the specific instability of the cauchy horizon with respect to the kink - mode perturbation was considered in the case of the reissner
nordstrm(anti ) de sitter black hole @xcite .
the problem of the cauchy horizon instability has been probed also by different numerical calculations ( see , e.g. , @xcite and references therein ) . the mass inflation problem in the case of the rotating black holes was elaborated in @xcite .
the standard mass inflation calculations @xcite are based on the using of the generalized dray thooft
redmount ( dtr ) relation @xcite in the linear approximation of the einstein equations near the perturbed inner horizon . however , the using of linear approximation to the dtr relation near horizons is a quite improper in view of the nonlinearity of the einstein equations .
this nonlinearity is especially crucial in the vicinity of the black hole horizons .
an additional weak point in the standard mass inflation calculations is in a fallacious using of the _ global _ null cauchy horizon as a place for the maximal growth of the back - reaction perturbations instead of the _ local _ inner apparent horizon , which is separated from the cauchy horizon .
the maximal back - reaction perturbations inside the charged black hole take place ( besides the central singularity ) at the local inner apparent horizon and not at the separated global cauchy horizon .
this qualitative feature was missed in the previous mass inflation calculations .
it is shown below that a back - reaction of two counter - streaming light - like fluxes result in only the small corrections to the background metric near the local inner apparent horizon of the charged black hole by taking into account its separation from the cauchy horizon .
this implies the absence of the mass inflation and stability of the charged black hole interiors .
we use a ( slow ) quasi - stationary approximation , when both the inflowing and outflowing radial energy fluxes of light - like particles are small , @xmath0 and @xmath1 . respectively ,
the rate of black hole mass growth is also small ( for more details see , e.g. , @xcite ) . in finding the back - reaction in this approximation
, we use the reissner
nordstrm black hole solution as a background metric and retain in equations only those perturbation terms , which are not higher , than the linear ones with respect to @xmath2 and @xmath3 .
in other words , we calculate the back - reaction in the linear perturbation approximation with respect to the small dimensionless energy flux parameters , @xmath4 and @xmath3 .
a general space - time metric in the spherically symmetric case can be written in the form @xcite : @xmath6 -r^2d\omega^2,\ ] ] where @xmath7 is a 2-sphere metric and @xmath8 and @xmath9 are two arbitrary functions of two coordinates @xmath10 and @xmath11 .
this is a coordinate frame of the _ eddington finkelstein _ type , related , in particular , with the ingoing null geodesics @xmath12 . in analogy with the schwarzschild , reissner
nordstrm and charged vaidya metric , we use the following form for the metric function @xmath9 : @xmath13 where @xmath14 is a mass function and @xmath15 is an electric charge . in the special case of @xmath16 ,
this metric is the charged vaidya solution . respectively , at @xmath16 and @xmath17 , this metric is the reissner
nordstrm solution . in the back - reaction calculations
, we follow very closely to @xcite , by using both the ingoing and outgoing vaidya metrics as perturbations for the background reissner nordstrm black hole metric . for the ingoing vaidya metric in the @xmath5-frame
we choose the mass function @xmath14 in the specific form : @xmath18 & \mbox { at } v\geq v_0 , \\
\!m_0(1-\beta_0 ) & \mbox { at } v < v_0 , \end{array } \right . \label{massin}\ ] ] with constants @xmath19 , @xmath20 , @xmath21 and @xmath22 .
respectively , for the corresponding outgoing vaidya solution in the @xmath23-frame we choose quite a similar expression for the mass function m(u , r ) : @xmath24 & \mbox { at } u\geq u_0 , \\
\!m_0(1-\beta_1 ) & \mbox { at } u <
u_0 , \end{array } \right . \label{massout}\ ] ] with the additional constants @xmath25 , @xmath26 and @xmath27 . the global geometry of the reissner nordstrm black hole , perturbed by both the ingoing and outgoing radial null fluxes , is shown in fig .
[ cpdiagram ] by means of the carter
penrose diagram . the metric functions @xmath28 and @xmath29 at both the inner and outer apparent horizons , @xmath30 and @xmath31 , of the perturbed black hole .
these local apparent horizons are the boundaries between the @xmath32- and @xmath33-regions with the different metric signatures , @xmath34 and @xmath35 , respectively .
the variable parts of these apparent horizons are shown by the thick curves @xmath36 , @xmath37 , @xmath38 and @xmath39 in fig . [ cpdiagram ] . , and outflowing , @xmath40 , radial fluxes of light - like particles , related with the corresponding ingoing and outgoing vaidya solutions outside the black hole . inside the black hole ,
i.e. , at @xmath41 and @xmath42 , there is a double filled overlapping region with two coexisting and oppositely directed fluxes . the global cauchy horizon @xmath43 , which is defined by the null line @xmath44 ,
is shown in part by the null line @xmath45 and further by the null dashed line @xmath46 .
the perturbed metric deviates from the linear sum of ingoing and outgoing vaidya metrics ( due to the nonlinearity of the einstein equations ) in the space - time regions , corresponding to the all future - directed light - cones in this overlapping region .
the maximal perturbation of the black hole metric , which is the most interesting for the discussed mass inflation problem .
the maximal blue - shift of the inward radiation , viewed by the free - moving observer , will take place just at the part of the local inner apparent horizon @xmath47 , shown by the thick time - like curve @xmath39 and not at the part of the global cauchy horizon @xmath43 , shown by the dashed null line @xmath48.,scaledwidth=80.0% ] there is only one flux , respectively , inflowing @xmath4 or outflowing @xmath49 in the filled but non - overlapping regions of the carter
penrose diagram in fig .
[ cpdiagram ] for the perturbed metric .
solutions of the einstein equations in these non - overlapping regions are the corresponding ingoing and outgoing charged vaidya metrics .
meanwhile , both the inward and outward fluxes coexist in the overlapping region .
it must be stressed , that the overlapping region exists only inside the black hole , at @xmath41 and @xmath42 . in the overlapping region
a corresponding solution of the einstein equations deviates from the vaidya metric . in other words ,
the sum of two counter - streaming vaidya solutions is not the valid solution of the einstein equations due to their nonlinearity . evidently , in the regions without any fluxes ( see the non filled regions in fig .
[ cpdiagram ] the standard reissner
nordstrm metric is realized , though with the different values of the black hole mass before and after the passage of inflowing or outflowing fluxes .
the metric functions @xmath28 and @xmath29 at both the inner and outer apparent horizons , @xmath30 and @xmath31 , of the perturbed black hole .
these local apparent horizons are the boundaries between the @xmath32- and @xmath33-regions with the different metric signatures , @xmath34 and @xmath35 , respectively .
a principal point is that the global cauchy horizon @xmath50 is separated from the local inner apparent horizon @xmath51 in the case of the non - stationary ( perturbed ) metric .
this separation is clearly viewed in fig .
[ cpdiagram ] , where the global cauchy horizon @xmath52 is shown in part by the null line @xmath45 and further by the dashed null line @xmath46 , while the corresponding part of the inner apparent horizon @xmath53 is shown by the time - like curve @xmath39 .
the maximal perturbation of the black hole metric , which is the most interesting for the discussed mass inflation problem , and the maximal blue - shift of inward radiation , viewed by the free - moving observer , will take place just at the @xmath47 ( the time - like curve @xmath39 ) and not at the @xmath43 ( the null line @xmath48 ) . the singular behavior of metric functions @xmath54 and @xmath55 at the horizons is especially crucial when the double - null ( u , v)-frame is used near horizons .
in particular , it was demonstrated the ill - posedness of a double null `` free '' evolution scheme in numerical calculations for the einstein equations , when constraints are imposed only at the boundaries , and all fields are propagated by means of the evolution equations @xcite .
this ill - posedness of the `` free '' evolution numerical scheme results in the artificial exponentially growing mode . to resolve a back - reaction problem for the perturbed spherically symmetric black hole ,
the two sought functions @xmath55 and @xmath54 in the general metric ( [ ef ] ) must be calculated by using the einstein equations with the appropriately chosen perturbations .
we follow below closely to e. poisson and w. israel @xcite by writing the einstein equations in the form @xmath56 where the maxwellian contribution to the energy - momentum tensor from the black hole electric charge @xmath15 is @xmath57 and @xmath58 is a perturbation energy - momentum tensor .
the nonzero components of the einstein tensor @xmath59 in the eddington
finkelstein frame ( [ ef ] ) are @xmath60 where dot @xmath61 , and prime @xmath62 .
the corresponding einstein equations are related by the bianchi identity . therefore , one of them is not independent , e.g. , equation , related with the component @xmath63 from ( [ eg22 ] ) . as a background metric
we consider the reissner
nordsrm black hole metric , which is an exact electro - vacuum solution of the einstein equations ( [ einstein ] ) with @xmath64 .
we calculate in the following the back - reaction on the background reissner nordsrm black hole metric of both the inflowing and outflowing radial fluxes of light - like particles , described by the perturbation energy - momentum tensor @xmath65 with , respectively , the energy influx @xmath66 and outflux @xmath67 and the radial null vectors @xmath68 where @xmath69 .
both the outer horizon @xmath70 and the inner horizon @xmath71 in the used quasi - stationary approximation ( with a linear accuracy in @xmath0 and @xmath72 ) are solutions of the equation @xmath73 or , formally , @xmath74 at @xmath75 .
note that these horizons are , respectively , the _ inner and outer apparent horizons _ for a non - stationary metric @xcite .
the perturbation energy - momentum tensor components in the @xmath5-frame ( [ ef ] ) are @xmath76 we calculate in the following the back - reaction solution of the einstein equations near horizons , at @xmath77 . in the quasi - stationary approximation
the both fluxes are nearly constant in the vicinity of horizons , @xmath78 and @xmath79 , respectively . from equation ( [ eg10 ] ) we get near horizons @xmath80 respectively , from equations ( [ eg00 ] ) and ( [ eg11 ] ) we get @xmath81 near horizons , where @xmath82 and @xmath83 , we have @xmath84 the mass function @xmath14 strongly depends on the coordinate @xmath11 but weakly on the coordinate @xmath10 near @xmath85 in the used quasi - stationary approximation . for this reason , it is credible to adopt a factorization for the mass function near horizons ( see also @xcite ) : @xmath86 with the dimensionless coordinate @xmath87 and with the `` mass '' @xmath88 , where @xmath89 is from equation ( [ massin ] ) .
the function @xmath90 is weakly growing with @xmath10 , i.e. , @xmath91 and , therefore , @xmath92 .
to solve the first nonlinear equation in ( [ mrv ] ) near the outer and inner apparent horizons @xmath93 , we define the black hole extreme parameter @xmath94 and the new variable @xmath95,\ ] ] the horizons @xmath96 are solutions of equation @xmath97 .
near horizons , at @xmath98 , we have @xmath99 now the left equation in ( [ mrv ] ) is written as @xmath100 solution of this equation in the non - overlapping region , where there is only influx , is @xmath101 . evidently , this solution coincides with the corresponding ingoing vaidya metric in the non - overlapping region in the vicinity of the null line @xmath102 in fig .
[ cpdiagram ] .
meantime , solution of the same equation ( [ mrvhor ] ) in the vicinity of the inner apparent horizon , at @xmath103 , in the overlapping region , where there are both the inward and outward fluxes , is @xmath104 the resulting metric deviates from the vaidya solution in the overlapping region near the inner apparent horizon @xmath105 ( in the vicinity of the curve @xmath102 in fig .
[ cpdiagram ] ) .
the integration constant in ( [ muxrn ] ) is @xmath106 due to continuity relation , i.e. , this solution must coincide with the vaidya solution at the inner horizon @xmath105 on the border between the overlapping and non - overlapping regions ( at the point @xmath107 in fig .
[ cpdiagram ] ) . at the same time , the solution of equation ( [ mrvhor ] ) near both the inner and outer apparent horizons , at @xmath93 , in the non - overlapping regions , where there is only outward flux @xmath3 is the outgoing vaidya metric : @xmath108 this outgoing vaidya solution , however , is presented here in the ingoing ( v , r)-frame , which is non - optimal in the presence of outward flux @xmath3 . in the outgoing
( u , r)-frame the mass function @xmath109 would have the vice - versa behavior .
in general , there are two variable branches of the inner apparent horizon , @xmath51 and @xmath53 , shown by the thick curves @xmath36 and @xmath39 in fig .
[ cpdiagram ] .
the back - reaction mass function @xmath110 from ( [ muxrn ] ) is finite at the both inner apparent horizons @xmath36 and @xmath39 without any indication of the mass inflation .
an alternative approach is in using of the general spherically - symmetric metric in the schwarzschild - like @xmath111-frame @xcite : @xmath112 with two arbitrary functions , @xmath113 and @xmath114 . for application to the back - reaction problem of the accreting matter on the charged black hole , we define , additionally , two metric functions , @xmath115 and @xmath116 or , equivalently , two _ mass functions _ , @xmath117 and @xmath118 : @xmath119 the nonzero components of the einstein tensor @xmath59 in the @xmath111-frame ( [ spherical ] ) have the following form @xcite : @xmath120 -\frac{e^{-\sigma}}{2 } \left[\eta''+(\eta'-\sigma')\left(\frac{\eta'}{2}+\frac{1}{x}\right)\right]\!.\end{aligned}\ ] ] for the inward and outward null vectors in the energy - momentum tensor ( [ tensorinout ] ) we choose now @xmath121 the corresponding perturbation energy - momentum tensor ( [ tensorinout ] ) is now @xmath122 from equations ( [ einstein ] ) , ( [ g01 ] ) , ( [ g00 ] ) and ( [ g00 ] ) with perturbation energy - momentum tensor from ( [ t00tr ] ) and ( [ t01 t ] ) we get the requested back reaction equations for the mass functions @xmath117 and @xmath118 in the vicinity of the apparent horizons at @xmath123 , where @xmath124 and @xmath125 : @xmath126 in the quasi - stationary approximation the both mass functions , @xmath117 and @xmath118 , strongly depend on the radial coordinate @xmath11 , but very weakly on the time coordinate @xmath127 . for this reason , as in the previous section [ sectionvr ] , we adopt near horizons a factorization for the mass functions : @xmath128 and @xmath129 with the dimensionless coordinate @xmath87 and with a weakly growing `` mass '' @xmath130 .
we use also the variable @xmath131 from equation ( [ deltapm ] ) to solve the nonlinear equation ( [ primem ] ) in the vicinity of the inner and outer apparent horizons . again , as in the section [ sectionvr ] , the outer and inner apparent horizons at @xmath96 are solutions of equation @xmath97 , where @xmath132 is defined by equation ( [ deltapm ] ) .
now , near the apparent horizons , at @xmath133 , where @xmath98 , we have @xmath134 respectively , equation ( [ primem ] ) is written near the apparent horizons as @xmath135 the resulting solutions for mass functions near the apparent horizons are @xmath136 here @xmath137 and the values of mass functions at the horizons , respectively , @xmath138 and @xmath139 are @xmath140 it can be seen in fig .
[ cpdiagram ] , that all the variable parts of the horizons , i.e. , the apparent horizons @xmath141 and @xmath142 , are placed in the non - overlapping regions of the carter
penrose diagram , where there is only one flux , inward or outward ( curves @xmath36 , @xmath39 , @xmath143 and @xmath38 ) . at the same time
, the constant parts of the inner and outer horizons are placed either at the borders of the non - overlapping regions with the zero fluxes ( null lines @xmath144 and @xmath145 ) , corresponding to the static black hole with a mass @xmath146 , or in the regions without any flux ( null lines @xmath147 and @xmath148 ) , corresponding to the static black holes with masses @xmath149 and @xmath150 .
note , that the null line @xmath145 , which is a constant part of the inner apparent horizon @xmath47 is identical to the part of cauchy horizon @xmath43 ) of the global metric .
the most interesting for us is the time - like curve @xmath39 of the inner apparent horizon @xmath47 of the perturbed black hole . the mass functions @xmath151 and @xmath152 are equal at @xmath53 with a corresponding value @xmath153 according to equation ( [ mu+2 ] ) . from equation ( [ murn0 ] ) for the black hole with the charge @xmath154 and flux @xmath155 near the inner apparent horizon @xmath105 , modified by the back - reaction .
inside the filled box the used linear perturbation approximation in @xmath156 is insufficient for the exact determination of the mass function @xmath157 . in the used approximation ,
the value of the @xmath157 at the @xmath105 is finite and defined by equation ( [ mu- ] ) .
perturbations of the mass functions @xmath157 and @xmath158 remain small at the inner apparent horizon , @xmath159.,scaledwidth=80.0% ] solution ( [ murn0 ] ) for the mass function @xmath157 near horizons may be written in the form of the inverse function @xmath160 , where @xmath161\!-\!1\!\right\}\!.\ ] ] here , the signs @xmath162 and @xmath163 are related to the different brunches of the function @xmath164 : the first branch for @xmath165 and the second one for @xmath166 .
these branches are shown in fig .
[ mu0minus ] for the case of @xmath157 near the inner apparent horizon , corresponding to the curve @xmath39 in fig .
[ cpdiagram ] , at @xmath103 .
the first branch is at @xmath165 and the second one is at @xmath166 . with the chosen approximations , solutions ( [ murn0 ] ) ,
( [ murn1 ] ) and ( [ inverse ] ) are valid only in the narrow region @xmath167\dot m < |\delta_\pm|\ll1 $ ] near the apparent horizons . in these solutions
we retain only the leading perturbation terms @xmath168 and neglect the much more smaller contributions , @xmath169 .
the used linear perturbation approximation with respect to the small parameter @xmath170 would be insufficient for calculations of the mass functions at @xmath171 , where the quadratic perturbation terms @xmath172 and the higher orders ones must be taking into account .
see in fig .
[ mu0minus ] the mass function @xmath157 from equation ( [ murn0 ] ) near the inner apparent horizon @xmath173 .
note also , that divergence in ( [ mu+2 ] ) in the extreme limit at @xmath174 is related with a general instability of the perturbed extreme black hole ( for some details see , e.g. , @xcite ) . the derived solutions ( [ murn0 ] ) and ( [ murn1 ] ) demonstrate that the back - reaction corrections to the mass functions @xmath117 and @xmath118 are small near and at the inner apparent horizon of the non extreme black hole .
namely , the relative disturbance of the inner apparent horizon is small , of the order of @xmath175 .
therefore , the mass inflation phenomenon is absent at least in the used quasi - stationary approximation .
we calculate the blue - shift of the inward radiation viewed by a free - moving observer , traversing the non - stationary local apparent horizon @xmath71 or the global cauchy horizon @xmath52 of the perturbed charged black hole .
the new feature in this calculation , with respect to a previous similar blue - shift calculation in @xcite , is in taking into account , at fist , the non - stationarity of the black hole due to the presence of the influx of light - like particles and , at second , the additional perturbation , produced by the observer himself .
the appropriate place for a possible large blue - shift is a border between the @xmath176-region and @xmath177-region inside the charged black hole . in the discussed model ,
this border composed of the part of the inner apparent horizon @xmath47 and the part of cauchy horizon @xmath43 , shown , respectively , by the time - like curve @xmath39 and the null line @xmath145 in fig .
[ cpdiagram ] .
following closely to @xcite , we use , at first , the ingoing vaidya solution as a background metric with a mass function @xmath90 from equation ( [ massin ] ) .
the influx of light - like particles in the ingoing vaidya metric is described by the energy - momentum tensor @xmath178 , where @xmath179 and @xmath180 .
the requested energy density , measured by a free - moving observer with four - velocity @xmath181 , is @xmath182 where @xmath183 may be calculated , e.g. , from ( [ massin ] ) . by using the four - velocity normalization @xmath184
, we get the trajectory equations for the free - moving observers of two dissimilar kinds ( starting for simplicity somewhere inside the @xmath176-region in fig .
[ cpdiagram ] ) : @xmath185,\ ] ] where overdot denotes differentiation with respect to the observer s proper time .
observer of the first - kind ( we call him the `` red - shift observer '' ) is traversing the border between the @xmath176-region and @xmath186-region , while the second - kind observer ( we call him the `` blue - shift observer '' ) is traversing the border between the @xmath176-region and @xmath177-region ( see fig .
[ cpdiagram ] ) . for the red - shift observer in the @xmath176-region and near the apparent horizon @xmath71 , where @xmath187 , @xmath188 or @xmath189 , we would have from ( [ v ] ) the trajectory equation @xmath190\approx-\frac{2}{\dot r}.\ ] ] it is evidently seen in fig .
[ cpdiagram ] that any red - shift observer traverses the non - stationary part of the inner apparent horizon @xmath71 , shown by the space - like curve @xmath36 , at the finite values of @xmath10 .
respectively , the value of @xmath191 is finite at @xmath192 according to equation ( [ v1 ] ) . in result
, the red - shift observer will see the finite gravitational red - shift of the influx @xmath4 at the inner apparent horizon @xmath51 .
meantime , for the blue - shift observer , in the same @xmath176-region and near the apparent or cauchy horizons , we would have from ( [ v ] ) quite the different trajectory equation @xmath193\approx\frac{2\dot r}{f}\ ] ] the blue - shift observer may traverse either the _ non - stationary _ part of the inner apparent horizon @xmath47 , shown by the time - like curve @xmath39 in fig .
[ cpdiagram ] , or the stationary part of the inner apparent horizon , which is the part of the global cauchy horizon , @xmath194 , shown by the null line @xmath145 in fig .
[ cpdiagram ] .
the inner apparent horizon @xmath47 is placed in the space - time beyond the null - line @xmath195 because @xmath196 . for this reason
we need to use the @xmath197-frame for the discussed ingoing vaidya solution to solve the trajectory equation for the blue - shift observer , traversing the inner apparent horizon @xmath47 . to circumvent this difficulty we note , that the blue - shift observer in the @xmath5-frame corresponds to the red - shift one in the @xmath23-frame .
this means that a blue - shift observer will see the finite gravitational red - shift of the outflux @xmath3 , traversing the inner apparent horizon @xmath53 , for the similar reasons , as the discussed earlier , the red - shift observer , traversing the inner apparent horizon @xmath51 . to calculate the values of the possible blue - shift or red - shift , viewed by observers traversing the inner apparent horizons @xmath71 , it is needed to use the geodesic equations in the perturbed vaidya metric , which is beyond the scope of this paper . for the sake of this paper
it is enough to establish that these values are finite .
quite the contrary , the blue - shift observer traverse the cauchy horizon @xmath43 at the infinite value of coordinate @xmath10 .
therefore , @xmath198 , is infinite at the cauchy horizon for these observers .
we integrate now equation ( [ v ] ) near the cauchy horizon @xmath43 ( i.e. , in the vicinity of the null line @xmath45 in fig .
[ cpdiagram ] ) .
differentiation of the metric function @xmath199 from ( [ f ] ) with respect to variables @xmath10 and @xmath11 with a mass function @xmath90 for the ingoing vaidya metric yields the trajectory equation for a free moving second kind observer in the @xmath176-region near the cauchy horizon : @xmath200 where @xmath201 is a surface gravity at the cauchy horizon @xmath202 . substituting @xmath203 from ( [ v2 ] ) into equation ( [ dfdv ] ) , we obtain @xmath204 it is crucial that a term with @xmath205 in ( [ dfdv2 ] ) , related with the black hole non - stationarity , was missed in previous calculations of the infinite blue - shift at the cauchy horizon .
solution of equation ( [ dfdv2 ] ) with @xmath88 from ( [ massin ] ) is @xmath206 where @xmath207 is an incomplete gamma - function .
the asymptotic form of this solution is @xmath208 we neglect in this expression the terms of the order of @xmath209 and higher . in result ,
@xmath210 at @xmath211 , indicating the power - law blue - shift divergence at the cauchy horizon at @xmath212 .
e. poisson and w. israel obtained a much more stronger exponential divergence for @xmath213 ( see , e.g. , eqs .
( b7 ) and ( b8 ) in @xcite ) by neglecting the black hole non - stationarity , i.e. , by putting @xmath214 in equation ( [ dfdv2 ] ) .
the derived blue - shift at the cauchy horizon @xmath50 is infinite for quite a formal reason : it was considered the static black hole metric with @xmath215 with the ignorance of the inevitable perturbation of the black hole metric , produced by the moving observer .
this perturbation , even if extremely small , will separate the local inner apparent horizon @xmath47 from the global cauchy horizon . physically , this metric perturbation from the free moving observer may be qualitatively modeled by the discussed outflux @xmath1 . in result , the blue - shift observer will traverse the local inner apparent horizon @xmath47 , viewing only the finite blue - shift of the influx .
note that the blue - shift observer will not see any considerable blue - shift of ingoing radiation @xmath4 at the dashed part of the cauchy horizon @xmath50 ( null line @xmath48 in fig .
[ cpdiagram ] ) .
solution of the perturbation back reaction problem for the non extreme reissner - nordstrm black hole reveals no indication of the mass inflation by taking into account the separation of the inner apparent horizon from the cauchy horizon in the ( slow ) quasi - stationary accretion approximation .
this separation was missed in the previous calculations of the mass inflation phenomenon .
the relative back - reaction corrections to the perturbed metric in the ( v , r)-frame and ( t , r)-frame at the finite distance from both the inner and outer apparent horizons , @xmath216 , appear to be of the order of small accretion rate , @xmath170 , which is a small dimensionless energy flux parameter .
at the same tine , near and at the apparent horizons , at @xmath123 , the relative back - reaction corrections to the black hole metric are the largest , but still remain the small , of the order of @xmath217 .
this means the absence of mass inflation phenomenon inside the charged black hole . additionally , it shown that a back reaction removes the infinite blue - shift singularity at the inner apparent horizon .
there are a lot of limitations to the validity of the derived back - reaction corrections .
the most vulnerable approximation is the slow accretion rate .
it must be stressed also that semi - classical effects may influence the behavior of matter in the vicinity of the cauchy horizon .
note also that the left patch of the carter - penrose diagram of the used eternal black hole geometry is absent in the general gravitational collapse .
a further clarification of the mass inflation problem is required beyond the quasi - stationary limit , e.g. , by using the detailed numerical calculations with the large accretion rate .
author acknowledges e.o .
babichev , v.a .
berezin and yu.n .
eroshenko for the enlightening discussions .
this work was supported in part by the the research program ofn-17 of the division of physics , russian academy of sciences .
99 poisson e and israel w 1989 1663 poisson e and israel w 1989 b * 233 * 74 poisson e and israel w 1990 d * 41 * 1796 hamilton a j s and avelino p p 2010 _ phys . rep .
_ * 495 * 1 ( arxiv:0811.1926 [ gr - qc ] ) hwang dong - il , lee bum - hoon and yeom dong - han 2011 _ jcap _ * 12 * 006 ( arxiv:1110.0928 [ gr - qc ] ) brady p r , nunez d and sinha s 1993 d * 47 * 4239 ( arxiv : gr - qc/9211026 ) maeda h , torii t and harada t 2005 064015 ( arxiv : gr - qc/0501042 ) hod s and piran t 1998 1554 ( arxiv : gr - qc/9803004 ) hod s and piran t 1998 gen .
. grav . * 30 * 1555 ( arxiv : gr - qc/9902008 ) hong s e , hwang d , stewart e d and yeom d 2010 045014 ( arxiv:0808.1709 [ gr - qc ] ) ori a 1992 2117 redmount i h 1985 _ prog .
phys . _ * 73
* 1401 dray t and thooft 1985 _ comm . math . phys . _ * 99 * 613 dokuchaev v i and eroshenko yu n 2011 d * 84 * 124022 ( arxiv:1107.3322 [ gr - qc ] ) babichev e , dokuchaev v and eroshenko yu 2012 115002 ( arxiv:1202.2836 [ gr - qc ] ) bondi h 1964 ( london ) a * 281 * 39 landau l d and lifshitz e m 1975 _ the classical theory of fields _ ( oxford : pergamon press ) gundlach c and pullin j 1997 991 ( arxiv : gr - qc/9606022 ) booth i and martin j 2010 d * 82 * , 124046 ( arxiv:1007.1642 [ gr - qc ] ) | the mass inflation phenomenon implies that black hole interiors are unstable due to a back - reaction divergence of the perturbed black hole mass function at the cauchy horizon .
weak point in the standard mass inflation calculations is in a fallacious using of the global cauchy horizon as a place for the maximal growth of the back - reaction perturbations instead of the local inner apparent horizon .
it is derived the new spherically symmetric back - reaction solution for two counter - streaming light - like fluxes near the inner apparent horizon of the charged black hole by taking into account its separation from the cauchy horizon . in this solution
the back - reaction perturbations of the background metric are truly the largest at the inner apparent horizon , but , nevertheless , remain small .
the back reaction , additionally , removes the infinite blue - shift singularity at the inner apparent horizon and at the cauchy horizon . |
hantaviruses are rodent - borne zoonotic agents that may cause diseases in humans such as hemorrhagic fever with renal syndrome and hantavirus pulmonary syndrome @xcite .
hantaviruses have been identified at an increasing rate in recent years , and as of now about thirty different ones have been discovered throughout the world .
one of these , the sin nombre virus , was not isolated until 1993 after an outbreak in the four corners region in the usa @xcite .
the host of this particularly dangerous virus is the deer mouse , _ peromyscus maniculatus _ , the most numerous mammal in north america .
the virus produces a chronic infection in the mouse population , but it is not lethal to them .
it is believed that transmission in the rodent population is horizontal and due to fights , and that the subsequent infection of humans , where the mortality rate can be as high as fifty percent , is produced by their contact with the excreta of infected mice .
moreover , so far there is no vaccine or effective drug to prevent or treat the hantavirus pulmonary syndrome .
therefore , a major effort has been made to understand the population dynamics of deer mice colonies in order to design effective prevention policies @xcite .
it has been noted that environmental conditions are directly connected to outbreaks of hanta @xcite .
for instance , the 1993 and 1998 outbreaks occurring in the four corners region have been associated with the so - called @xmath0 @xmath1 southern oscillation @xcite .
related to this and other such observations , the effects of seasonality in ecological systems have been a subject of recent interest @xcite .
multi - year oscillations of mammal populations @xcite , prey - predator seasonal dynamics @xcite , and persistence of parasites in plants between seasons @xcite are examples that illustrate the importance of seasonality in population dynamics .
recently , abramson and kenkre have proposed a phenomenological model for mice population that successfully reproduces some features of hantavirus propagation @xcite . in particular
, that model explores the relation between resources in the medium , carrying capacity , and the spread of the infection in the rodent colony .
herein we study the effects of seasonality in that model .
our motivation is not only to provide more realism to the model , but also to investigate the counterintuitive effects that dynamic alternation may cause in a biological system .
brownian motors @xcite and switching - induced morphogenesis @xcite are examples that show that alternation in time of `` uninteresting '' dynamics may produce `` interesting '' outcomes . along these lines
, we will show that alternation of seasons , neither of which by itself fulfills the environmental requirements on the carrying capacity for spreading of the infection , may produce an outbreak of the disease .
the mechanism driving this behavior is the interruption of the relaxation process that equilibrates the mouse population from season to season : if the duration of seasons becomes shorter than the relaxation time of the population , the disease spreads .
the paper is organized as follows . in sec .
[ sec2 ] we briefly review the model for mouse populations introduced in @xcite . in sec .
[ sec3 ] we explain how seasonality is introduced in that model and analyze the conditions for hanta outbreaks to take place due to the alternation of seasons . the exact solution of the model and a particular example that illustrates the phenomenology is given in sec .
[ sec4 ] . finally , in sec.[sec5 ] , we summarize the main conclusions and propose some directions for future work .
the model introduced in @xcite for the temporal evolution of a population of mice subjected to the hantavirus infection reads : @xmath2 where @xmath3 and @xmath4 are the population densities of susceptible and infected mice respectively , @xmath5 is the total population of mice , and @xmath6 , @xmath7 , @xmath8 and @xmath9 are positive constants .
the terms on the right hand sides of eqs .
( [ model1 ] ) and ( [ model2 ] ) take into account the following processes : births with rate constant @xmath7 , depletion by death with rate constant @xmath8 , competition for the resources in the medium characterized by the carrying capacity @xmath9 , and transmission of the infection with rate coefficient @xmath6 .
it is worth noting that infected pregnant mice produce hanta antibodies that keep their fetus free from the infection ; that is , _ all _ mice are born susceptible @xcite , as indicated by the absence of a birth term in eq .
( [ model2 ] ) .
note also the absence of a recovery term in the model since , as mentioned earlier , mice become chronically infected by the virus . the system of equations ( [ model1 ] ) and ( [ model2 ] ) has four equilibrium points .
two of them are irrelevant for the analysis : the null state @xmath10 , which is always unstable if @xmath11 ( a condition that we will assume throughout this paper ) , and a meaningless state with @xmath12 for any value of the parameters .
the other two equilibria are @xmath13 @xmath14 the stability of the equilibrium points ( [ equilibrio1 ] ) and ( [ equilibrio2 ] ) depends on the value of the carrying capacity . if @xmath15 then ( [ equilibrio1 ] ) is stable and ( [ equilibrio2 ] ) unstable . if @xmath16 it is the other way around .
that is , when the available resources , @xmath9 , are below the critical value , @xmath17 , the infection does not propagate in the colony , the whole population of mice grows healthy , and its size increases proportionally with those resources .
as soon as @xmath9 surpasses @xmath17 the virus spreads in the colony , the susceptible mouse population saturates , and the fraction of infected mice becomes larger as @xmath9 increases ( see fig .
[ fig1 ] ) .
the four corners region , where an important number of cases of hantavirus pulmonary syndrome have occurred , has a desert climate .
the largest climate variations within this region come from the alternation between dry and rainy seasons .
we will therefore assume alternation in time of these two seasons .
it is important to remark that a two - season assumption is not crucial , and that the analysis with four seasons is also straightforward within the formalism introduced herein . during each of the two seasons under consideration
we assume there to be no climate variations , so that each season can be characterized by a set of time - independent parameters @xmath18 where @xmath19 .
we implement square - periodic season alternation where the duration of each season is @xmath20 . again
, this particular alternation pattern is not essential for the mechanism .
other schemes of season alternation , e.g. different duration of the seasons or random switching between seasons , do not qualitatively change the phenomenology .
any quantity @xmath21 alternating in the way described above can be written as : @xmath22 where @xmath23 and @xmath24 is a periodic square wave @xmath25 let us now suppose the following conditions for the sets @xmath26 according to seasonality : @xmath27 where 1 stands for the rainy season and 2 for the dry one .
the biological motivation for these conditions is the following .
the dry season provides less resources for the colony than the rainy season @xmath28 , and as a consequence the death rate is higher @xmath29 , and the transmission rate is also larger since fights for the available resources are expected to increase @xmath30 .
however , notice that we consider the birth rate to be larger during the dry season @xmath31 .
there are two reasons for this .
first is the assumption that the colony makes an attempt to maintain its population .
second is an implementation of the maturation process in the model : baby mice do not contribute to the propagation of the disease @xcite , and , therefore , even if births were actually more numerous during the rainy season , the contribution of this fact to the propagation of the infection will only be important in the next ( dry ) season , when mice have matured and are ready to fight .
it will be shown later that these assumptions lead to a situation with high population during the rainy season and low population during the dry one , in agreement with the available data @xcite .
moreover , we assume that @xmath32 , i.e. , the resources are _ all times _ ( during both seasons ) _ below _ the minimum critical threshold that triggers the propagation of the disease .
we will show that nevertheless it is possible for the infection to spread .
the equilibrium populations of the susceptible and infected mice are determined by the set of values @xmath26 .
when switching from season to season , the populations evolve trying to reach a new equilibrium .
therefore , the dynamics is driven by the competition between two characteristic times .
on the one hand there is an _ external _ time scale determined by the seasonal forcing , @xmath33 .
on the other hand , the relaxation toward equilibrium after a switching of seasons involves a relaxation time . the latter measures the time required for the mouse colony to relax to the equilibrium state associated with @xmath34 after having been driven during the previous season by the conditions @xmath26 , that is , @xmath35 where @xmath36 .
the _ internal _ time scale is defined as the fastest relaxation process , i.e. , @xmath37 $ ] .
`` fast '' or `` slow '' seasonal alternation then refers to the comparison between these two time scales . if @xmath38 the mouse population has enough time to accommodate to the new conditions from season to season and relax to equilibrium .
moreover , since we have imposed the condition that the resources at any time of the year are below the critical thresholds @xmath39 , there will be no infected mice . in the other limit , @xmath40 , seasonal changes occur too fast , the relaxation process is interrupted , and no equilibrium can be reached from season to season .
then note that an adiabatic elimination can be implemented @xcite , and @xmath41 in eq .
( [ alternation ] ) can be replaced by its average value , @xmath42 .
therefore , in the limit of fast season alternation the system is driven by the set of averaged values @xmath43 , and the critical carrying capacity is given by @xmath44 .
as a consequence , it is possible to find regions of parameters where @xmath45 is smaller than the _ effective _ value of the carrying capacity associated with the averaged values : @xmath46 ^{-1}=\left[\frac{1}{2}\left(\frac{1}{k_{1}}+\frac{1}{k_{2}}\right)\right]^{-1}=\frac{2k_{1}k_{2}}{k_{1}+k_{2}},\ ] ] and the infection propagates .
general conditions leading to this behavior can be posed , but the expressions are rather cumbersome .
we prefer , for the sake of simplicity , to show a particular typical case .
we use the following values for the parameters : @xmath47 @xmath48 that lead to @xmath49 and @xmath50 respectively .
the dynamics are completely determined once the value of the carrying capacity during each season is specified . according to the previous discussion
, these parameters can be chosen such that the following conditions hold : @xmath51 ^{-1}>k_{c+ } , \quad k_{1}>k_{2 } , \quad k_1<\min(k_{c1},k_{c2}).\ ] ] these conditions lead to the points @xmath52 that fulfill the seasonal requirements given by eq .
( [ conditions ] ) , so that slow alternation of seasons leads to infection - free states while fast alternation leads to hanta outbreaks .
this region is plotted in fig .
notice in particular that the point @xmath53 lies inside the region and that @xmath54 . in the next section
we illustrate the seasonality - induced propagation of the disease for this particular point .
so far we have determined that outbreaks of hanta induced entirely by seasonal changes can occur if the duration of the seasons are short enough .
now we establish the meaning of `` short enough '' quantitatively .
since @xmath45 is strictly smaller than the effective value of the carrying capacity , there should be a finite value of @xmath55 such that for any @xmath56 the population of infected mice is greater than zero , but for periods above this critical period the infected population goes to zero . in order to obtain the value of the critical period we need to solve the system of equations ( [ model1 ] ) and ( [ model2 ] ) . in spite of its nonlinearities
the system can be solved analytically by means of a reciprocal transformation @xcite and the following exact solution is obtained : @xmath57 where @xmath58 and @xmath59 are the initial conditions for @xmath4 and @xmath3 respectively , and the following definitions have been introduced , @xmath60 @xmath61 because the external forcing due to the alternation of seasons is periodic , floquet theory ensures the existence of a periodic solution .
the values of @xmath58 and @xmath59 compatible with the non - equilibrium periodic solution can be obtained by evolving the system during the first half of a period under dynamics @xmath62 and the second half under dynamics @xmath63 , and forcing periodity on the solutions after a whole period of evolution , that is , @xmath64 in order to close the system in the non - equilibrium steady state @xmath65 and @xmath66 , the values of @xmath58 and @xmath59 that solve that system of equations ( [ algebra1 ] ) and ( [ algebra2 ] ) must be then re - introduced in eqs .
( [ solution1 ] ) and ( [ solution2 ] ) .
the critical period is then the largest value of @xmath67 satisfying the condition @xmath68 we illustrate the procedure with the example mentioned above where the parameters are given by eqs .
( [ parameters1 ] ) and ( [ parameters2 ] ) , and with @xmath69 and @xmath70 . the results are shown in figs .
[ fig3 ] and [ fig4 ] .
the values of @xmath71 and @xmath72 as a function of the period of the seasons are depicted in fig .
[ fig3 ] , where the populations of the susceptible and infected mice at the end of each season are given . as seen in that figure ,
the value of the critical period is @xmath73 .
note that if the alternation is slow , @xmath74 , all mice grow healthy . on the other hand , if the alternation is faster than the relaxation time required by the colony to accommodate its population from season to season , @xmath75 , the virus spreads and @xmath76 .
notice that in the limit @xmath77 the dynamics is driven by @xmath78 and the populations of susceptible and infected mice are given by eq .
( [ equilibrio2 ] ) with @xmath79 , @xmath80 , @xmath81 , and @xmath82 .
let us stress again that the carrying capacity is below its critical threshold at any time . in fig .
[ fig4 ] we plot , for different period lengths , the exact solutions @xmath65 and @xmath66 as a function of time through one period of evolution .
the first semi - period corresponds to the rainy season and the second to the dry season . when seasons last long ( left panel ) , there are no infected mice and the susceptible population simply oscillates between the two equilibrium states given by eq .
( [ equilibrio1 ] ) . for sufficiently short seasons ( right panel ) , there is propagation of the disease and the values of @xmath71 and @xmath72 fluctuate around the equilibrium points determined by eq .
( [ equilibrio2 ] ) and the set of parameters @xmath78 . finally ,
when the period of the seasons is near , but below , the critical period ( central panel ) , the infected population is small and the population @xmath4 oscillates in a more pronounced manner .
by introducing seasonality in a paradigmatic model for hantavirus propagation in mice colonies , we have shown that the alternation of seasons may cause outbreaks of the disease .
the striking feature of that behavior lays in the fact that neither season satisfies the conditions for the infection to spread in terms of the availability of resources .
the mechanism responsible for the phenomenon is the competition between two time scales : an external one , the duration of a season , and an internal one , the relaxation time for the mouse colony to equilibrate its population from season to season .
we have shown that if the duration of the seasons is longer than the relaxation time , no propagation of hantavirus occurs .
on the other hand , if the relaxation process is interrupted by a fast seasonal alternation , the disease spreads .
we have analyzed the general conditions for which the phenomenon occurrs .
moreover , we have illustrated the mechanism with a particular example that can be solved exactly .
this work may help to clarify the reported relation between climate and propagation of hanta in mice populations . however , to elucidate whether the proposed phenomenon actually takes place in nature we depend on data that unfortunately are not available in the literature .
one can envision further modifications of the model that may improve its features , such as , for example , the inclusion of spatial dependence or of noisy contributions to the dynamics .
finally , we stress that the general idea underlying the mechanism is model - insensitive and can therefore be extended to other systems where seasonality plays a relevant role .
work along these directions is in progress .
the authors gratefully acknowledge fruitful comments from and discussions with v. m. kenkre during the elaboration of this work . c. escudero is grateful to the department of chemistry and biochemistry of the university of california , san diego for its hospitality . this work has been partially supported by the national science foundation under grant phy-9970699 , by the ministerio de educacin y cultura ( spain ) through grants no .
ap2001 - 2598 and ex2001 - 02880680 , and by the ministerio de ciencia y tecnologa ( spain ) , project
bfm2001 - 0291 .
99 j. n. mills , t. l. yales , t. g. ksiazek , c. j. peters and j. e. childs , emerg .
dis . * 5 * , 95 ( 1999 ) .
d. m. engelthaler , d. g. mosley , j. e. cheek , c. e.levy , k. k. komatsu , p. ettestad , t. davis , d. t. tanda , l. miller , j. w. frampton , r. porter and r. t. bryan , ibid . * 5 * , 87 ( 1999 ) .
j. n. mills , t. g. ksiazek , c. j. peters and j. e. childs , ibid . * 5 * , 135 ( 1999 ) .
a. j. kuenzi , m. l. morrison , d. e. swann , p. c. hardy and g. t. downard , ibid . * 5 * , 113 ( 1999 ) .
k. d. abbott , t. g. ksiazek and j. n. mills , ibid .
* 5 * , 102 ( 1999 ) . c. h. calisher , w. sweeney , j. n. mills and b. j. beaty , ibid .
* 5 * , 126 ( 1999 ) . v. w. kenkre , _ memory formalism , nonlinear techniques , and kinetic equation approaches _ , in _ modern challenges in statistical mechanics : patterns , noise , and the interplay of nonlinearity and complexity _ , aip conference proceedings * 658 * , 2003 . | using a model for rodent population dynamics , we study outbreaks of hantavirus infection induced by the alternation of seasons .
neither season by itself satisfies the environmental requirements for propagation of the disease .
this result can be explained in terms of the seasonal interruption of the relaxation process of the mouse population toward equilibrium , and may shed light on the reported connection between climate variations and outbreaks of the disease . |
one of the greatest of the many mysteries of radio pulsars is their evolutionary histories . as pulsars age
, powerful electromagnetic torques act to increase the rotation period , @xmath3 , of the strongly magnetized neutron star , causing the spin - frequency , @xmath4 , to decrease over time , a phenomenon known as spin - down .
this can be described by the equation @xmath5 where @xmath6 is a parameter known as the braking index , with @xmath7 for a magnetic dipole rotating in a vacuum .
the spin - down torque acting on the star is proportional to @xmath8 , and it decays over time as @xmath8 decreases . there is also evidence that @xmath9 is not constant , but rather is also changing over time . the surface dipole magnetic field strength at the magnetic equator , assuming that the spin and magnetic axes are orthogonal , is conventionally given by @xmath10 ; at the magnetic poles the field strength is @xmath11 @xcite .
@xmath12 tends to be bigger for younger pulsars than for older ones having a large spin - down age , @xmath13 $ ] .
this in turn suggests that @xmath9 tends to reduce with increasing age , and much conjecture remains about the origin(s ) of this evolution .
many authors have argued that the main contributor to the @xmath9-evolution is the progressive alignment of the spin and magnetic axes : e.g. @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , @xcite and @xcite .
but @xcite and @xcite have argued in favour of a random distribution of the inclination angles between the angular velocity and magnetic axes and against magnetic alignment .
others have favoured magnetic field decay ( e.g. * ? ? ?
* ) , while there are also authors who argue for continuing counter - alignment to eventual orthogonality of the axes , via an electromagnetic torque exerted by magnetospheric electric currents flowing on open field lines @xcite .
according to @xcite , pulsars initially counter - align ( via a strongly temperature - dependent dissipative torque in the fluid interior ) to reach the orthogonal rotator state after about @xmath14yr , where they remain for @xmath15@xmath16yr ( while the dissipative torque decays as the interior cools ) , after which alignment ( via electromagnetic torque ) occurs .
if @xmath9 were constant , the braking index @xmath6 would be equal to the _ apparent _ braking index , @xmath17 if @xmath18 can be determined from observations of a pulsar , then @xmath19 can be computed . however , if @xmath9 is varying , then in general @xmath20 . some authors ( eg .
* ; * ? ? ?
* ) have argued that , for pulsars of moderate age , @xmath19 exceeds the value of 3 that corresponds to magnetic dipole radiation .
this could be taken to imply that either magnetic alignment or field decay must be occurring .
however , it is likely that recovery from unseen pulsar glitches is the principal contribution to measured variations in @xmath21 @xcite for these pulsars . in this paper
, we shall use the pulsar evolution models from ( * ? ? ?
* hereafter cb83 ) , ( * ? ? ?
* hereafter cb86 ) and ( * ? ? ?
* hereafter c93 ) , which all invoke a progressive alignment of the spin and magnetic axes .
the cb83 model develops the pulsewidth - age relation on the assumption that all pulsars move along the same evolutionary track , differing only in age and orientation with respect to earth .
the cb86 model relaxes this assumption and allows a distribution of alignment time - scales , as well as a time - varying relationship between characteristic and actual age .
the c93 models additionally allow a distribution of initial inclination angles .
all of these models predicted a minimum in the mean pulsewidth as a function of characteristic age and provided a good fit to the available data , consisting of 293 pulsewidth values from the @xcite catalogue .
here we use the atnf ( australia telescope national facility ) pulsar catalogue @xcite , version 1.35 , to investigate pulsar evolution through magnetic alignment using the candy blair models .
this contemporary catalogue contains much larger data sets of 872 and 1420 isolated radio pulsars with 10 per cent and 50 per cent intensity pulsewidth values , @xmath22 and @xmath23 .
it also contains many more old pulsars than was previously the case , which tend to have lower radio luminosities .
it thus allows a more precise analysis of pulsar evolution . as a consistency check
, we also examine separately those pulsars with pulsewidths measured during the parkes multibeam pulsar survey @xcite , giving @xmath22 and @xmath23 data sets of 377 and 934 pulsars respectively .
the paper is structured as follows .
section [ sec : magneticalignmenttheory ] presents a summary of the theory of alignment between the spin and magnetic axes ; section [ sec : pulsewidthevolution ] gives a mathematical description of pulsewidth evolution ; in section [ sec : data : fitting ] we least - squares fit the evolutionary pulsewidth models to the @xmath22 and @xmath23 data as functions of characteristic age ; in section [ sec : analysis ] we further analyse these results , in particular comparing them to the inclination angle estimates of @xcite and @xcite ; and conclusions are presented in section [ sec : conclusions ] .
the candy blair pulsar evolution models are based on two effects .
one is progressive alignment of the spin and magnetic axes , for which the formula describing the alignment phase of the @xcite model is adopted : this has an exponential decay of the sine of the magnetic - spin inclination angle , @xmath24 , from its initial value @xmath25 : @xmath26 where @xmath27 is the pulsar s age and @xmath28 is the alignment time - scale .
the alignment has a simple electromechanical origin : the electromagnetic radiation emitted by an oblique rotating dipole results in a torque on the star that causes the angular velocity axis to migrate through the neutron star toward alignment with the magnetic axis .
the second effect is progressive narrowing of the emission cone , described by a power - law dependence of its half - angle , @xmath29 , on the rotation period , @xmath3 : @xmath30 with @xmath31 a positive constant having a value between @xmath32 and @xmath33 @xcite . for a pulsar with constant rotational inertia and fixed magnetic moment vector , the rotation period evolves according to @xmath34 = constant
. it was pointed out long ago @xcite that this simple rule is incompatible with the observed distribution of pulsars in the @xmath3@xmath21 plane , conflicting with stationary ` flow ' through that plane to pulsar ` death ' at advanced ages .
however , alignment produces an effective reduction in the magnetic moment , reducing the torque in proportion to @xmath35 ; so , in the jones model , the period evolves according to @xcite @xmath36 with @xmath6 ( the braking index ) constant .
equation ( [ eq : pdotpn ] ) integrates to give : @xmath37 where @xmath38 is the pulsar s initial period and @xmath39 is its limiting period as @xmath40 . assuming that pulsars eventually slow substantially from their initial spin period , @xmath41
, we can approximate equation ( [ eq : p : t ] ) by @xmath42^{1/(n-1 ) } \ , .
\label{eq : p : t : approx}\ ] ] using ( [ eq : p : t : approx ] ) for @xmath43 in ( [ eq : rho : t ] ) for @xmath44 yields a simple approximation for the evolution of @xmath29 , expressed directly in terms of @xmath27 : @xmath45^{-\gamma /(n-1 ) } \ , , \label{eq : rho : t : approx}\ ] ] where @xmath46 is the limiting emission cone half - angle as @xmath47 .
we note that , because @xmath38 has effectively been taken as zero in going from ( [ eq : p : t ] ) to ( [ eq : p : t : approx ] ) for @xmath43 , equation ( [ eq : rho : t : approx ] ) is not applicable to the very youngest pulsars : for @xmath48 it reduces to @xmath49 which diverges as @xmath50 . for the typical parameter values ( see section [ sec : datafits ] table [ table :
parameters : three : ngamma : min ] ) @xmath51 @xmath52 @xmath53yr , and @xmath54 , equation ( [ eq : rho : t : approx:2 ] ) gives @xmath55 for @xmath56yr . as only one known pulsar is this young , the divergence is outside the age range we can analyse , and the approximation ( [ eq : rho : t : approx ] ) for @xmath44 is valid for our purposes .
a pulsar s spin - down age , @xmath13 $ ] , depends on @xmath6 , which is going to be a variable in our fitting procedure below . instead of the spin - down age , from here on
we use the the characteristic age which is defined to be @xmath57 the characteristic age is independent of @xmath6 and corresponds to the spin - down age in the case @xmath7 .
its use enables us to show fittings for different @xmath6 on one plot using the same characteristic - age data . inserting equation ( [ eq : p : t : approx ] ) for @xmath43
, we find that @xmath58 is related to the pulsar s actual age , @xmath27 , by : @xmath59[\,\exp(2t/\tau ) - 1\ , ] \label{eq : tc : t}\\ \mbox{or~~ } \exp[2t(t_c)/\tau ] & = & 1 + ( 4t_{c})/[(n-1)\tau ] \mbox{~. } \label{eq : t : tc}\end{aligned}\ ] ] note that factors of @xmath60 effectively rescale @xmath61 to @xmath62 in these equations .
it follows from equation ( [ eq : tc : t ] ) that @xmath63 for relatively young pulsars ( young compared with the spin - down time - scale ) ; for old pulsars , @xmath64 increases as an exponential function of @xmath65 .
the characteristic and actual ages coincide for relatively young pulsars in the dipole rotator case , for which @xmath66 .
the @xmath58@xmath27 relation enables equations ( [ eq : alphat ] ) , ( [ eq : p : t : approx ] ) and ( [ eq : rho : t : approx ] ) for @xmath24 , @xmath3 and @xmath29 to be re - expressed in terms of @xmath58 which has the advantage of being directly determined from the measured quantities @xmath3 and @xmath21 instead of in terms of @xmath27 : @xmath67^{-1/2 } \mbox{~ , } \label{eq : alpha : tc } \\
p(t_c ) \ ! & = & \ ! p_{\infty}\left[1+{(n-1)\tau}/{4t_c } \right]^{-1/(n-1 ) } \mbox{~ , } \label{eq : p : tc } \\ \mbox{and}\hspace{5 mm } \rho(t_c ) \ ! & = & \ !
\rho_{\infty } \left[1+{(n-1)\tau}/{4t_c } \right]^{\gamma/(n-1 ) } \mbox{~. } \label{eq : rho : tc}\end{aligned}\ ] ] equation ( [ eq : rho : tc ] ) describes an emission - cone half - width that decreases quite rapidly with time for young pulsars ; consequently , the observed pulsewidths should decrease with age for the youngest pulsars .
the alignment of the spin and magnetic axes , described by equation ( [ eq : alpha : tc ] ) , should cause an increase in the observed pulsewidths for older pulsars , as the pulse will occupy a larger fraction of a complete rotation .
the combination of these two processes should lead to a minimum in the observed angular pulsewidths for pulsars of moderate ages ( cb83 ) .
note that taking the derivative of equation ( [ eq : p : t ] ) and using equation ( [ eq : characteristicage ] ) for @xmath58 gives ( cf .
equation ( [ eq : brakingindex : definition ] ) ) @xmath68 showing that the apparent braking index increases with characteristic age .
equations ( [ eq : alpha : tc ] ) , ( [ eq : p : tc ] ) and ( [ eq : rho : tc ] ) expressing the evolution of the magnetic - spin inclination angle , the pulsar period and the emission cone half - angle in terms of a pulsar s characteristic age can conveniently be re - written in terms of a normalized dimensionless characteristic age : @xmath69\ ; , \label{eq : tc : defn}\ ] ] which reduces to @xmath64 in the dipole rotator @xmath70 case .
thus : @xmath71^{-1/(n-1 ) } \mbox{~ , } \label{eq : p : tc } \\
\mbox{and~~ } \rho(t_c ) \ ! & = & \ !
\rho_{\infty}[1+(2t_c)^{-1}]^{\gamma/(n-1 ) } \mbox{~. } \label{eq : rho : tc}\end{aligned}\ ] ] in the equations used for the pulsewidth modelling below , namely ( [ eq : alpha : tc ] ) and ( [ eq : rho : tc ] ) , or ( [ eq : alpha : tc ] ) and ( [ eq : rho : tc ] ) , for @xmath72 and @xmath73 , @xmath6 appears only in the two combinations @xmath74 and @xmath75 .
hence there are only four basic independent parameters in the candy - blair alignment models .
later we will set @xmath76 , consistent with the @xmath31 values found by other authors ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
hence , it is convenient in this study to define the four independent parameters to be @xmath25 , @xmath46 , @xmath77 and @xmath78 , where @xmath79 since these last two parameters are normalized to the @xmath76 case . in terms of these parameters ,
the two combinations @xmath74 and @xmath75 become @xmath80 \mbox{~,~and}\\ ( n-1 ) \,\tau & = & ( n_{\gamma}-1)\tau_{\gamma } \mbox{~ , } \label{eq : parameters : gamma : tau:2}\end{aligned}\ ] ] and equation ( [ eq : napp : n : tc ] ) gives @xmath81
for an individual pulsar , the observed angular pulsewidth is given as a function of @xmath82 and @xmath83 by @xmath84 where @xmath83 is the angle between the observer s direction and the pulsar s spin axis @xcite .
the strong dependence on the orientation of the observer with respect to the spin axis leads to a wide scatter in the pulsewidth values , so we need to take a mean pulsewidth for pulsars with certain values of @xmath24 and @xmath29 . the mean pulsewidth is obtained by integrating ( [ eq : w : defn ] ) over the angular extent of the emission beam : @xmath85 in which @xmath86 is the probability that the angle between the spin axis and the observer s direction is in the range @xmath83 to @xmath87 .
the integration is over all angles @xmath83 for which emission is both directed toward the observer and seen as pulsed .
the limits of integration in ( [ eq : w : bar ] ) for @xmath88 are explained by means of fig .
[ fig:1 ] . for a pulsar that is not too close to alignment , such that the emission beam does not contain the rotation axis ( fig .
[ fig:1 ] , left ) , @xmath89 and the beam is intercepted for a part ( but not all ) of the rotation period if @xmath83 is in the range @xmath90 to @xmath91 . for a pulsar that is close to alignment , with the emission beam containing the rotation axis ( fig .
[ fig:1 ] , right ) , @xmath92 and the range @xmath93 is excluded , as some part of the beam is then always directed toward the observer , assuming the emission cone to be filled .
we take @xmath94 ; the choice of first or second quadrant for @xmath24 is a matter of convention , switching it between hemispheres being equivalent to reversing the sense of the neutron star s rotation . the beam , however , can extend into the other hemisphere , which it will do if @xmath95 , as expected for young pulsars with wide beams . if a pulsar produces observable radio emission from both poles , then this will be observed as a main pulse and interpulse , provided that the pulsar is nearly orthogonal or the pulsar beams are wide .
however , the mean pulsewidths referred to here are for emission from only one of these poles , not the combined main pulse interpulse width . either greater ( left ) or less ( right ) than the emission cone half - angle @xmath29 . in the latter case , the emission cone encloses the spin axis , and there may be little or no modulation ; in particular , to observers in the cone @xmath96 , the emission cone is continuously visible .
the left - hand diagram depicts the usual case of a pulsar with moderate values of @xmath24 and @xmath29 , typical of mature - aged pulsars .
the right - hand diagram depicts a pulsar with a thin beam and nearly aligned spin and magnetic axes , typical of old pulsars in the candy - blair theory and the jones model.,width=317 ] on taking pulsar spin axes to be randomly directed with respect to observers , and normalizing over the angular extent of the emission cone , the probability distribution for @xmath83 becomes : @xmath97 where @xmath98 is the ` beaming fraction ' , which measures the fraction of the sky swept out by the beam .
equation ( [ eq : prob : zeta ] ) is the ratio of the surface area of a ring on a sphere , corresponding to observers in the range @xmath83 to @xmath99 , to the surface area of the sphere swept out by the beam .
the surface area of the part of a sphere between co - latitudes @xmath100 and @xmath101 , normalized to the surface area of a sphere , is @xmath102 \sin\left[(\theta_2 - \theta_1)/2\right ] \mbox{~. } \label{eq : sphere : area}\ ] ] inserting @xmath103 and @xmath104 gives the beaming fraction : @xmath105 applicable to both situations depicted in fig .
[ fig:1 ] .
we note that ( * ? ? ?
( 7 ) ) use a different beaming fraction for the case @xmath92 , namely @xmath106 , which results from integration over 0 to @xmath107 .
as they remark , there may be little or no modulation when @xmath92 , and we prefer to exclude the ` continuously seen ' cone , @xmath108 ( fig .
[ fig:1 ] , right ) . the evolutionary time dependence of @xmath24 and @xmath29 , equations ( [ eq : alphat ] ) and ( [ eq : rho : t : approx ] ) , results in a mean pulsewidth , @xmath109 , that is a function of time only , @xmath110 .
this function is to be treated as a density , @xmath111 giving the mean pulsewidth of pulsars aged between @xmath27 and @xmath112 .
the age of the pulsars will here be measured in terms of their log characteristic age , @xmath113 , and will be binned with respect to that variable . under the change of variables @xmath114 , we have @xmath115 the factor @xmath116 is given by @xmath117 where the @xmath27@xmath61 relation , equation ( [ eq : t : tc ] ) , has been used .
so @xmath116 increases monotonically from @xmath118 for young pulsars , flattening to @xmath119 for old ones . to allow for variations between the evolutionary histories of different pulsars , cb86 and c93 relaxed the assumptions that all pulsars are born with the same inclination angle and have the same alignment time - scale , @xmath120 and @xmath28 respectively . instead ,
these are replaced by a probability distribution function , @xmath121 , for initial inclination angles ( see section [ sec : alignmentmodels ] ) , and a log - normal distribution of alignment time - scales : @xmath122 \mbox { , } \label{eq : prob : log : tau}\ ] ] where @xmath123 and @xmath124 are the mean and standard deviation of the logarithm of the time - scale distribution , @xmath125 . if the normalized parameters @xmath77 and @xmath78 that were introduced in equations ( [ eq : parameters : gamma : n])([eq : parameters : gamma : tau:2 ] ) are used , then @xmath126 has a log - normal distribution with mean @xmath127 and standard deviation @xmath124 unchanged .
the mean observed pulsewidth as a function of characteristic age is now given by @xmath128 in which the normalization factor , @xmath129 , is @xmath130 the beaming fraction , @xmath131 , must be included as a weighting factor in equation ( [ eq : wbar : tc ] ) to take into account the effect of beaming on the probability of observation , and those weights are normalized by the factor @xmath132 ( the mean pulsewidth is only computed for those pulsars that are observed ) .
the aim here is to fit equation ( [ eq : wbar : tc ] ) for @xmath133 to the observed data . in appendix
[ sec : appendix : rrb ] it is shown that @xmath134\pi & \mbox{for } & \alpha \geq \rho\mbox{~ , } \\
\left[\cos(\rho-\alpha)-\cos\rho\right]\pi & \mbox{for } & \alpha \leq \rho \mbox{~. } \end{array } \right .
\label{eq : pwint}\ ] ] inserting this result into equation ( [ eq : wbar : tc ] ) for @xmath133 allows a considerable improvement in the numerical performance of the curve - fitting . at a given characteristic age ,
equation ( [ eq : rho : tc ] ) implies that @xmath135 if a pulsar were to have @xmath136 , where @xmath137 \label{eq : tau : max } \\ & \approx & \frac{4 t_{c}}{n-1 } \left ( \frac{90\degr}{\rho_{\infty } } \right)^{\frac{n-1}{\gamma } } \mbox{,~~for~ } \rho_{\infty }
\ll 90\degr \mbox{.}\end{aligned}\ ] ] this unphysical possibility arises from the approximation made earlier that @xmath138 . in reality
a pulsar with @xmath139 must be born with a characteristic age greater than the @xmath61 used in equation ( [ eq : tau : max ] ) .
therefore integrals over @xmath28 are terminated at an upper limit of @xmath140 , where the factor 0.99 is included to assist numerical convergence .
we use three different initial alignment models , @xmath121 , from c93 that involve different distributions of the initial inclination angles .
model i assumes that all pulsars start with their spin and magnetic axes mutually perpendicular : @xmath141 .
this corresponds to the @xcite model following the initial brief phase of counter - alignment to orthogonality .
model ii uses the probability distribution @xmath142 which assumes pulsars are born with a random distribution of inclination angles from 0 to @xmath143 .
model iii uses @xmath144 which assumes pulsars tend to be born with inclination angles closer to @xmath145 than would occur by chance . at the outset
there is no particular reason to think that model iii might be approximately correct ; it is included here for contrast with models i and ii .
both 10 per cent and 50 per cent intensity pulsewidth data , @xmath22 and @xmath23 , are used in this analysis of pulsewidth evolution .
the data were retrieved from the atnf pulsar catalogue @xcite , which includes results from several surveys at around 400 and 1400 mhz . for this study , all the binary , millisecond , anomalous x - ray , rotating radio transient ( rrat ) and globular cluster pulsars have been removed from the data set , because these pulsars are considered to follow different evolutionary histories from those of the more common isolated radio pulsars . the condition for excluding millisecond ( or low magnetic field ) pulsars
was chosen to be @xmath146 .
pulsars with interpulses whose @xmath22 or @xmath23 values potentially include both the main pulse and interpulse were also removed . to do this the complete census of the interpulse pulsar population recently published by @xcite was used .
it identified 27 pulsars with interpulses , and our fig .
[ fig : ip : histrogram ] shows histograms of the existing @xmath22 and @xmath23 values for them : it is evident that the distributions can be treated as bimodal with one maximum well below @xmath143 and the other at around @xmath147 .
for the purposes of this study it was assumed that a value of @xmath148 includes emission from both poles of the pulsar ( that is both the main pulse and the interpulse ) , and so should not be included in the study .
this resulted in 6 pulsars being removed from the @xmath22 data set leaving a net total of 872 pulsars ; and 3 interpulse pulsars being removed from the @xmath23 data set , leaving a net total of 1420 pulsars .
these two data sets are referred to hereafter as ` atnf cat @xmath22 ' and ` atnf cat @xmath23 ' .
( black ) and @xmath23 ( grey ) pulsewidths for the interpulse pulsars listed by @xcite ; the bars are stacked for clarity .
it is evident that the data distributions can be treated as bimodal .
widths greater than @xmath149 were assumed to have included the interpulses and were excluded from the @xmath22 and @xmath23 pulsewidth data sets used in this paper.,width=317 ] in a mean pulse profile the spacing between components usually increases , and with it the whole profile width , as the radio frequency , @xmath150 , decreases .
for instance , by measuring the profile widths of 6 pulsars up to 32 ghz , @xcite found that @xmath151 where @xmath152 , @xmath153 and @xmath154 are constant for each pulsar , with @xmath155 .
the @xmath22 and @xmath23 pulsewidths tabulated in the atnf catalogue are not all measured at the same observing frequency .
the frequency dependence of the pulsewidth therefore introduces some statistical noise into the ` atnf cat @xmath22 ' and ` atnf cat @xmath23 ' data sets . as a consistency check ,
we have also examined the subsets consisting of only those pulsars whose catalogued pulsewidth values were measured in the parkes multibeam survey at 1374 mhz @xcite .
these subsets are referred to hereafter as ` parkes mb @xmath22 ' and ` parkes mb @xmath23 ' , and contain 377 and 934 pulsars respectively . both the @xmath22 and @xmath23 characteristic age
pulsewidth data distributions naturally lend themselves to binning in @xmath113 , and it was also found numerically optimal to fit the mean pulsewidth , equation ( [ eq : wbar : tc ] ) , as a function of @xmath113 rather than @xmath61 , i.e. as @xmath133 . the @xmath22 or @xmath23 data
set is then binned to give @xmath156 data points , @xmath157 , @xmath158 .
it is desirable to make the bins as narrow as possible in order to give maximum resolution in @xmath159 .
this is because if a bin @xmath160 extends over a large range of @xmath159 in a region where the underlying function , here @xmath161 , has a significant slope , then a part of the uncertainty in the estimated value of the mean pulsewidth , @xmath162 , will result from the change in the underlying @xmath133 across the bin .
on the other hand , it is desirable to include as many points as possible in a bin to reduce the uncertainty in the @xmath162 estimate of @xmath163 $ ] , and also to ensure that the errors in the estimates are approximately normally distributed so that standard data - analysis procedures can be applied . to attempt to balance these competing effects
we binned the data so that each bin contained about 30 pulsars
the minimum number of points to ensure that the errors are approximately normally distributed .
least - squares fittings are made to both the @xmath22 and @xmath23 data sets .
even though there are more data points in the @xmath23 than the @xmath22 data sets , the @xmath22 is superior for our pulsewidth fitting because the @xmath23 set suffers more from confusion .
this is because a @xmath23 measurement will include outlying components in a profile only if they are above 50 per cent of the maximum intensity .
a @xmath22 measurement is more likely to include the outlying components and so will give a more consistent estimate of the total pulsewidth .
a histogram of the ratio of @xmath164 is found to be unimodal , with a peak at about 2 .
however , it extends to ratios above 10 , demonstrating that the @xmath22 in those cases is detecting outlying emission missed by the @xmath23 measurement .
the trade - off in relying on @xmath22 values is that the @xmath22 measurement is more difficult to make than the @xmath23 , and so there are fewer pulsars with existing @xmath22 measurements .
the data sets considered in this paper are summarized in table [ table : datasets ] . for each data
set it gives the name of the set and the number of pulsars it contains , and describes the binning of the data .
[ fig : w10:w50:data : comparison ] shows the atnf cat data sets and compares them to the corresponding parkes mb data sets .
we see that the atnf cat and parkes mb data sets have very similar distributions with respect to characteristic age .
.summary of the four data sets considered in this paper . [ cols="<,>,<",options="header " , ] [ table : alignment : timescales ] from equation ( [ eq : mu : gamma : log ] ) , the fitted value of @xmath123 is determined by the fitted value of @xmath165 and the value of @xmath31 : @xmath166 as plotted in fig . [
fig : mulog ] .
we see that a mean alignment time - scale @xmath167yr requires @xmath168 ; that is inconsistent with previous findings that @xmath31 is between @xmath169 and @xmath33 .
conversely , for that range of @xmath31-values , the curves in fig .
[ fig : mulog ] give a mean alignment time - scale of around @xmath170yr .
): the mean log alignment time - scale , @xmath123 , as a function of the parameter @xmath31 , which was introduced in equation ( [ eq : rho : t ] ) .
the solid curve corresponds to @xmath171 from the atnf cat @xmath22 model ii fit ( table [ table : parameters : three : ngamma : min ] ) , and the dashed curves are three standard deviations of error away from the black curve , @xmath172 .
previous studies have found a log alignment time - scale @xmath173 .
this is inconsistent here with the empirical finding that @xmath174.,width=317 ] caution is required in using the characteristic age as a measure of the true ages of pulsars , as will now be explored . defining @xmath175 and using equations ( [ eq : parameters : gamma : tau ] ) and
( [ eq : parameters : gamma : tau:2 ] ) involving @xmath78 , it follows from equation ( [ eq : t : tc ] ) for @xmath176 that @xmath177 \mbox{~. } \label{eq : tgamma : tc}\ ] ] the mean log age , @xmath178 , as a function of the log characteristic age , @xmath113 , is found by integrating over the distribution functions @xmath121 and @xmath179 . consider the best fit to the atnf cat @xmath22 data having parameters as listed in table [ table : parameters : three : ngamma : min ] .
a plot of @xmath180 for model ii is shown in fig . [
fig : logt ] ; it shows that that @xmath178 and @xmath113 agree up to @xmath181 , after which time @xmath178 starts to fall below @xmath113 .
the curves for models i and iii are similar . , ( solid curve ) and log age , @xmath182 , ( broken curves ) as functions of log characteristic age for model ii : ( dashed ) @xmath183 , ( dot - dashed ) @xmath184 ; and ( dotted ) @xmath185 .
the parameters are from the best fit to the atnf cat @xmath22 data set ( table [ table : parameters : three : ngamma : min ] ) , and the straight line is @xmath186.,width=317 ] however , the mean value @xmath187 applies to the population as a whole . for individual pulsars , the plots of @xmath188
depend on the value of @xmath78 for that pulsar .
various tracks for individual pulsars are shown in fig .
[ fig : logt ] : it is evident that in general the characteristic age diverges much more rapidly from the true age than for the populations mean ages , with the divergence being more rapid the smaller the alignment time - scale .
recall , however , that pulsars with shorter alignment time - scales have a smaller beaming fraction than those with longer time - scales , and so are less likely to be observed .
this highlights the uncertainty that may be present in previous studies that have used the characteristic age as a measure of true age when attempting to measure the alignment time - scales .
the characteristic age can over - estimate the true age , and so will tend to give larger alignment time - scales if it is the time measure .
from fig . [ fig : models:123 ] , showing mean pulsewidths versus @xmath189 , we have seen that the pulsewidth data clearly favour pulsar evolution through magnetic alignment , and not magnetic field decay alone nor progressive counter - alignment . there is a clear minimum of mean pulsewidth at about @xmath190yr .
the alignment process causes a significant increase in the mean @xmath22 pulsewidth for all characteristic age bins greater than @xmath191yr .
this is particularly significant in view of the fact that the atnf catalogue contains numerous pulsars with @xmath192yr , which were not available to the earlier studies .
the three models of initial alignment that were specified in section [ sec : alignmentmodels ] can not be tested by fitting the candy - blair models to the pulsewidth data alone . however , once a candy - blair model has been fitted to the pulsewidth data , it can be used to _ predict _ mean inclination angle evolution , @xmath193 . comparing these predictions to the datasets of @xcite and @xcite
, it was found that only model ii ( a random initial inclination angle ) gives a good fit to the data , as shown in fig .
[ fig : alpha : data ] .
the alignment time - scale found from fitting model ii to the atnf cat @xmath22 data ( see tables [ table : datasets ] and [ table : parameters : three : ngamma : min ] ) is @xmath194yr , which is at least a factor of ten shorter than the timescales determined by a number of other authors ( see table [ table : alignment : timescales ] ) .
the fit requires that the log alignment time - scale varies with a standard deviation of about @xmath195 across the observed population .
the fit gives a limiting cone opening half - angle of @xmath196 , and a braking index parameter @xmath77 of 2.30 , consistent with the mean , @xmath197 , of the measured apparent braking indices ( equations ( [ eq : brakingindex : definition ] ) and ( [ eq : napp : ngamma : tc ] ) , and table [ table : n ] ) .
consistent with empirical observations , the fit assumes the parameter @xmath76 ( see equation ( [ eq : rho : t ] ) ) . for other values of @xmath31
the fitted parameters @xmath6 and @xmath123 are determined from equations ( [ eq : parameters : gamma : n ] ) and ( [ eq : mu : gamma : log ] ) respectively . a plot of the mean beaming fraction for this fit is given in fig .
[ fig : fb : mean : scaled ] .
the initial predictions of pulsar alignment due to electromagnetic braking @xcite predicted time - scales of around @xmath198@xmath199yr , which was rather small .
the time - scales will be increased if pulsars are born with their magnetic and rotation axes nearly orthogonal .
@xcite @xcite found that the inclusion of a decreasing dissipative torque initially brought all pulsars close to counter - alignment , from which time the normal electromagnetic decay of @xmath24 would begin .
@xcite predicted that the alignment phase started at around @xmath198@xmath199yr , with a time - scale of around @xmath170yr , which was believed to be more in accord with the observed population .
we have found alignment time - scales consistent with the predictions of @xcite , though with a significant spread in values , down to the time - scales consistent with the absence of the counter - alignment phase .
contrary to the model of jones , though , we do not find evidence for all pulsars starting their alignment phase with @xmath120 close to @xmath143 ( i.e. model i ) .
this interesting counter - point to the model of jones deserves further investigation .
we thank the australia telescope national facility for making their pulsar database available .
we thank r. n. manchester for many helpful discussions and comments on the manuscript , and the referee of an earlier version of this paper for a helpful review .
beskin v. s. , gurevich a. v. , istomin y. n. , 1984 , ap&ss , 102 , 301 camilo f. , kaspi v. m. , lyne a. g. , manchester r. n. , bell j. f. , damico n. , mckay n. p. f. , crawford f. , 2000 , apj , 541 , 367 b. n. , 1993 , phd thesis , university of western australia candy b. n. , blair d. g. , 1983 , mnras , 205 , 281 b. n. , blair d. g. , 1986 , apj , 307 , 535 l. , goldstein m. , 1970 , apj , 159 , l81 gil j. a. , han j. l. , 1996 , apj , 458 , 265 gould d. m. , 1994 , phd thesis , the university of manchester gunn j. e. , ostriker j. p. , 1970 ,
apj , 160 , 979 johnston s. , galloway d. , 1999 , mnras , 306 , l50 p. b. , 1976a , nature , 262 , 120 p. b. , 1976b , apj , 209 , 602 jones p. b. , 1976 , ap&ss , 45 , 369 a. d. , wu x. , 1992 , ap&ss , 190 , 209 m. a. , kaspi v. m. , gavriil f. p. , 2005 ,
apj , 633 , 1095 m. a. , kaspi v. m. , gavriil f. p. , manchester r. n. , 2005 , apj , 619 , 1046 m. a. , kaspi v. m. , gavriil f. p. , manchester r. n. , gotthelf e. v. g. , kuiper l. , 2007 , ap&ss , 308 , 317 m. a. , kaspi v. m. , gotthelf e. v. , kuiper l. , 2006 , apj , 647 , 1286 d. r. , faulkner a. j. , lyne a. g. , manchester r. n. , kramer m. , mclaughlin m. a. , hobbs g. , possenti a. , stairs i. h. , camilo f. , burgay m. , damico n. , corongiu a. , crawford f. , 2006 , mnras , 372 , 777 lyne a. g. , manchester r. n. , 1988 , mnras , 234 , 477 lyne a. g. , pritchard r. s. , graham - smith f. , camilo f. , 1996 , nature , 381 , 497 lyne a. g. , pritchard r. s. , smith f. g. , 1993 , mnras , 265 , 1003 lyne a. g. , smith f. g. , 2006 , pulsar astronomy , 3rd ed .. cambridge university press , cambridge manchester r. n. , han j. l. , qiao g. j. , 1998 , mnras , 295 , 280 r. n. , hobbs g. b. , teoh a. , hobbs m. , 2005 , aj , 129 , 1993 manchester r. n. , lyne a. g. , camilo f. , bell j. f. , kaspi v. m. , damico n. , mckay n. p. f. , crawford f. , stairs i. h. , possenti a. , morris d. j. , sheppard d. c. , 2001 , mnras , 328 , 17 manchester r. n. , taylor j. h. , 1977 , pulsars .
freeman , san francisco manchester r. n. , taylor j. h. , 1981 , aj , 86 , 1953 m. m. , 1993 , apj , 413 , 317 f. c. , goldwire jr .
h. c. , 1970 , astrophys .
lett . , 5 , 21 narayan r. , ostriker j. p. , 1990
, apj , 352 , 222 u. s. , prasad s. s. , 1996 , a&a , 308 , 507 phinney e. s. , blandford r. d. , 1981 , mnras , 194 , 137 press w. h. , teukolsky s. a. , vetterling w. t. , flannery b. p. , 1992
, numerical recipes : the art of scientific computing , 2@xmath200 edition . cambridge university press , cambridge rankin j. m. , 1993a , apj , 405 , 285 rankin j. m. , 1993b , apjs , 85 , 145 ruderman m. a. , sutherland p. g. , 1975 , apj , 196 , 51 t. m. , konar s. , 2001 , a&a , 376 , 543 tauris t. m. , manchester r. n. , 1998 , mnras , 298 , 625 wang n. , manchester r. n. , zhang j. , wu x. j. , yusup a. , lyne a. g. , cheng k. s. , chen m. z. , 2001 , mnras , 328 , 855 p. , johnston s. , 2008 , mnras , 387 , 1755 xilouris k. m. , kramer m. , jessner a. , wielebinski r. , timofeev m. , 1996 , a&a , 309 , 481 w. , wu x. , 1991 , apj , 380 , 550
the integral that averages the pulsewidth over a random distribution of viewing angles can be evaluated analytically , as follows .
as above , @xmath24 is the pulsar magnetic - spin inclination angle , @xmath29 the emission cone half - width , and @xmath83 is the angle between the observer s direction and the pulsar s spin axis . in considering the variation of @xmath202 with viewing angle , it is helpful to employ the variable @xmath203 .
equation ( [ eq : a1 ] ) takes the form @xmath204
\,\mbox { , } \label{eq : a2}\ ] ] with the abbreviations @xmath205 note that @xmath206 the function @xmath207 , which is just the cosine of the pulse angular half - width , is @xmath208 throughout the cut across the emission beam by the sight line to any ` illuminated ' observer .
it reaches @xmath209 in magnitude ( i.e. @xmath210 ) at the pulse visibility limits @xmath211 substituting ( [ eq : a6 ] ) into ( [ eq : a3 ] ) gives @xmath212^{1/2 } } \;\mbox{. } \label{eq : a8}\ ] ] this shows that at @xmath213 , the ` outer ' visibility limit , @xmath150 is always positive , i.e. @xmath214 . at @xmath215 , the ` inner ' visibility limit , @xmath150 is @xmath214 for @xmath89 , @xmath145 for @xmath216 , and @xmath217 for @xmath218 .
let s visualize these three cases in turn .
\(1 ) for @xmath89 , the visibility limits @xmath219 both correspond to zero pulsewidth , @xmath220 .
equations ( [ eq : a4 ] ) and ( [ eq : a5 ] ) show that the minimum value @xmath221 of @xmath207 corresponding to the greatest pulsewidth occurs for the cone of observers at @xmath222 : @xmath223^{1/2 } \right\ } \label{eq : a9}\ ] ] at @xmath224 .
\(2 ) for @xmath216 , one can visualize the emission cone as rolling around the spin axis , which is then the inner visibility limit .
in this case , @xmath225^{1/2}$ ] ; also @xmath226 ( observer looking along the spin axis ) so @xmath227 , corresponding to @xmath228 .
so @xmath229 increases monotonically from @xmath145 at the outer visibility limit to @xmath230 on the spin axis in this special case .
\(3 ) for @xmath231 , the beam will be ` always on ' ( unpulsed ) to observers within @xmath232 of the spin axis .
so the pulse visibility limit @xmath215 then corresponds to @xmath233 , with @xmath207 ranging from @xmath214 at @xmath213 to @xmath217 at @xmath215 , with no minimum : @xmath229 increases monotonically from @xmath145 at the outer visibility limit to @xmath234 at the inner one .
integrating equation ( [ eq : a1 ] ) for @xmath235 over a random distribution of viewing angles , using the observer distribution function @xmath236 , gives the mean angular pulsewidth , for pulsars of given @xmath24 and @xmath29 : @xmath237\,\mathrm{d}x \;.
\label{eq : a11}\end{aligned}\ ] ] it is integration of the observer distribution function @xmath238 over the same range that yields the beaming factor @xmath239 . note for use below that @xmath240^{1/2 } & = & g^{1/2}(x)/(1-x^2)^{1/2 } \;\mbox { , } \label{eq : a12 } \\
\mbox{with~~ } g(x ) & { \ensuremath{\equiv } } & ( x - x_-)(x_+-x ) \;\mbox{. } \label{eq : a13}\end{aligned}\ ] ] the function @xmath241 describes an inverted parabola : it is zero at the limits of integration @xmath219 and positive in between , with a maximum value @xmath242 at @xmath243 .
applying integration by parts in equation ( [ eq : a11 ] ) for @xmath244 reduces the problem to one of evaluating an integral of an algebraic function . using @xmath245 together with equations ( [ eq : a4 ] ) and
( [ eq : a12 ] ) for @xmath246 and @xmath247 gives @xmath248 \nonumber \\ & & { } + { } \int_{x_-}^{x_+ } \frac{x(bx - a)\;\mathrm{d}x}{(1-x^2)g^{1/2}(x ) } \;\mbox { , } \label{eq : a14}\end{aligned}\ ] ] because the integrated part @xmath249 $ ] which by equation ( [ eq : a2 ] ) is just @xmath250 always vanishes at the limit @xmath213 , where @xmath251 always .
the first term on the rhs of equation ( [ eq : a14 ] ) is @xmath252 \mbox{~,}\\ \pi/2 & \mbox{for } & \alpha = \rho & [ f(x_+ ) = 0\;\mbox{,~}x_+ = + 1 ] \mbox{~,}\\ \pi\cos(\rho - \alpha ) & \mbox{for } & \alpha < \rho & [ f(x_+ ) = -1 ] \mbox{~. } \end{array } \label{eq : a15}\ ] ] this result shows that we can employ the following indefinite integrals , in which @xmath254 and @xmath255 are defined by equations ( [ eq : a7 ] ) and ( [ eq : a13 ] ) : @xmath256 { \mbox{~and } } \label{eq : a17}\ ] ] @xmath257^{1/2 } } } \nonumber \\ & = & \frac{2}{|\,a \pm b\,| } \arctan \sqrt \frac{(x - x_-)(1 \pm x_+)}{(x_+ - x)(1 \pm x_- ) } \;\;\mbox { , } \label{eq : a18}\end{aligned}\ ] ] where @xmath258 and equation ( [ eq : a7 ] ) for @xmath254 have been used .
so the required definite integrals are @xmath259 thanks to these definite integrals , equation ( [ eq : a16 ] ) for the observer - averaged pulsewidth evaluates to @xmath260 \;\mbox{. } \label{eq : a21}\ ] ] that is : @xmath261\pi & \mbox{for } & \alpha\geq\rho \;\mbox{,}\\ \left[\ , \cos(\rho - \alpha ) - \cos\rho \,\right]\pi & \mbox{for } & \alpha\leq\rho \;\mbox{,}\\ \end{array } \right .
\label{eq : a22}\end{aligned}\ ] ] as used in the main text
equation ( [ eq : pwint ] ) above . | using pulsewidth data for 872 isolated radio pulsars we test the hypothesis that pulsars evolve through a progressive narrowing of the emission cone combined with progressive alignment of the spin and magnetic axes .
the new data provide strong evidence for the alignment over a time - scale of about @xmath0myr with a log standard deviation of around 0.8 across the observed population . this time - scale is shorter than the time - scale of about @xmath1myr found by previous authors , but the log standard deviation is larger .
the results are inconsistent with models based on magnetic field decay alone or monotonic counter - alignment to orthogonal rotation .
the best fits are obtained for a braking index parameter , @xmath2 , consistent the mean of the six measured values , but based on a much larger sample of young pulsars . the least - squares
fitted models are used to predict the mean inclination angle between the spin and magnetic axes as a function of log characteristic age . comparing these predictions to existing estimates
it is found that the model in which pulsars are born with a random angle of inclination gives the best fit to the data .
plots of the mean beaming fraction as a function of characteristic age are presented using the best - fitting model parameters .
[ firstpage ] pulsars : general stars : evolution stars : magnetic fields stars : neutron . |
a large ion collider experiment ( alice ) uses a time projection chamber ( tpc ) @xcite as the central tracking detector for charged particle tracking .
currently the readout rate of the tpc is limited to 3.5 khz due to the use of triggered gating grid ( gg ) .
to cope up with the expected event rate of 50 khz for pb - pb collisions in the lhc run 3 after the long shutdown 2 ( ls2 ) , a major upgrade program for the alice tpc has been taken up @xcite . during this upgrade
, the multi wire proportional chambers ( mwpc ) at the end cap of the tpc will be replaced with detecting elements based on gas electron multiplier ( gem ) technology to restrict the ion back flow ( ibf ) into the drift volume to a tolerable limit in the absence of the gg .
the basic working principle of gem detector is the electron multiplication that happens when primary electrons pass through small holes in the gem foils where a very high electric field is imposed @xcite .
however , the electric field distribution inside the holes depends on the shape and size of the holes as well as the thickness of the polyimide foil . in consequence , the local variation of these parameters over the entire area of the foil results in variation of the gain . due to
the complications in the fabrication processes of gem foils , deviations from the design parameters over the entire area of the foils are not rare , specially when large sized foils are fabricated .
such variation of gain as a function of the hole diameter has been reported in a recent work @xcite .
but it is not possible to measure the gain for each individual foil before they are used to fabricate the detectors , when a large number of foils are involved .
so an independent evaluation of the variation of gain as a function of the physical parameters from simulation is necessary for better operation of the detector .
we have used a widely used commercially available finite element tool , ansys @xcite to create the electric field distribution inside the detector and open source garfield++ @xcite package to simulate the electron avalanche through the holes .
a more detailed description of the methodologies and results are discussed in section [ sim ] and section [ res ] respectively .
the detecting element of gem detector is a 50 @xmath0 m polyimide foil sandwiched between 5 @xmath0 m layers of copper on both sides . for foils produced with double mask technology ,
the holes have an hour - glass like shape where the diameter of the outer openings have typical values of 70 @xmath0 m and the diameter of the constricted part is 50 @xmath0 m .
single mask technology produces holes with conical shape with typical hole diameters of 70 @xmath0 m and 60 @xmath0 m on either sides of the cone .
a high electric field inside the hole is created and controlled by applying voltages on the metal layers .
electron multiplication happens when the electric field in the hole reaches a typical value of 10 kv / cm which is the limit of avalanche formation for the gas .
the geometry of the hole , material properties of the gem , the voltage applied to the metal layers and the boundary conditions are defined in ansys , which calculates the electric field mesh using finite element technique .
the overall formation of the electron avalanche ( fig .
[ fig_avalanche ] ) is simulated using garfield++ .
this program has interfaces with heed @xcite , which calculates the energy loss and ionisation of the gas molecules by moving charged particles and magboltz@xcite , which computes the transport properties of the electrons subjected to the electric field . finally for every single primary electron , the number of secondaries that cross the hole and reach certain distance below the bottom of the foil is counted . a distribution of this number for many thousands of avalanches is shown in fig .
[ fig_electron_distribution ] .
the mean of this distribution gives the gain .
electron multiplication across a typical double mask gem hole .
both electrons ( yellow ) and ions ( brown ) are shown . ]
distribution of the number of transfer electrons for a typical hole configuration of 70 - 50 - 70 @xmath0 m for top - middle - bottom hole diameter of a double mask gem . ]
it is important to mention here that the charging up effect of the gem foils has not been considered in this study .
we have taken five values of each of the three hole diameters to calculate the gain for bi - conical holes , on the other hand five values of each of the two hole diameters have been taken for conical holes .
100 k avalanches have been simulated for each setting .
@xmath1 ( 70:30 ) gas mixture has been used for all the calculations .
the values of the drift field ( @xmath2 ) and those of the induction field ( @xmath3 ) have been taken as 400 v / cm and 1 kv / cm respectively . a potential difference ( @xmath4 ) of 267
v has been applied across the foil .
the variation of the gain as a function of different hole diameters for the double and the single mask gem foils are shown in fig .
[ fig_2d_gain ] and fig .
[ fig_2d_gain_single_mask ] respectively . gain as a function of the hole diameters at the bottom and in the middle , keeping the diameter at the top fixed at 70 @xmath0 m . ]
gain as a function of the hole diameters at the bottom and that at the top for single mask gem foil . ]
we have studied the variation of the gain of single gem detector as a function of the shape of the holes using garfield++ simulation package .
the results reveal that this variation may reach up to 17 - 18% for a variation of 5 @xmath0 m in the hole diameter .
it is understood that the charging up of the gem foils in course of time might have effect on the results .
however these findings are needed to be compared with the measured data for similar variation .
the gain variation will be studied by changing other parameters such as gas mixture and @xmath4 as a continuation of this work .
the results of this study will help in qa of the gem foils where a large number of them is required such as in alice tpc .
this work was partially supported by the research grant sr / mf / ps-01/2014-bi from department of science and technology , govt . of india . c. garabatos , nucl .
instr . and meth .
a535 ( 2004 ) 197 . t. gunji , nucl . instr . and meth . a931 ( 2014 ) 1152
. f. sauli , nucl .
instr . and meth .
a386 ( 1997 ) 531 .
t. hildn _
et al . _ ,
instr . and meth .
a770 ( 2015 ) 113 .
http://www.ansys.com/ h. schindler and r. veenhof , http://garfieldpp.web.cern.ch/garfieldpp/ i. smirnov , http://consult.cern.ch/writeup/heed/ s. biagi , http://consult.cern.ch/writeup/magboltz/ | the alice experiment at lhc has planned to upgrade the tpc by replacing the mwpc with gem based detecting elements to restrict the ibf to a tolerable value . however
the variation of the gain as a function of physical parameters of industrially produced large size gem foils is needed to be studied as a part of the qa procedure for the detector .
the size of the electron avalanche and consequently the gain for gem based detectors depend on the electric field distribution inside the holes .
geometry of a hole plays an important role in defining the electric field inside it . in this work
we have studied the variation of the gain as a function of the hole diameters using garfield++ simulation package .
alice , time projection chamber , gas electron multipler |
wireless body area networks ( wbans ) represent the next generation of personal area networks @xcite .
the elementary components of a typical wban are sensors and a gateway device , which is alternatively known as a hub .
a traditional wban has a centralized topology , which is coordinated by the central hub .
it has a round - robin direct data exchange behavior between each sensor and gateway node @xcite .
this kind of system is widely used for patients monitoring and athletes training @xcite .
there were approximately 11 million active units around the word in 2009 , and this number is predicted to reach 420 million by 2014 @xcite .
the pervasive use of wban thus increases the need for good coexistence between multiple wbans .
imagine patients wearing such a system in a medical centre , the number of wbans closely located is large in some periods of the day and they can move rapidly with respect to each other .
due to wbans nature of high mobility and potential large density , it is generally not feasible to have a global coordinator in such a circumstance .
therefore , this leads to a challenging issue - interference between multiple closely located wbans .
interference among wbans is a major critical issue that can cause performance degradation and hence is a threat to reliable operation of any wban .
a proposed method that will be investigated in this paper is to use two - hop cooperative communication with an opportunistic relaying scheme .
recently , two - hop cooperative communications , which is included as an option in the ieee 802.15.6 standard @xcite , has been proved to overcome typical significant path loss experienced in single - link star topology wban communication @xcite .
several such cooperative communication schemes have been investigated for wbans using either narrow - band @xcite or ultra - wideband @xcite .
its effectiveness in interference mitigation has also been studied using realistic on - body channel data and a simulated model of inter - wban channel data @xcite .
it has been shown in @xcite that the use of cooperative communication in any wban - of - interest is able to significantly mitigate interference by providing an up - to 12 db improvement in sinr outage probability . in this paper , the work is extended from @xcite , some similarities are : 1 .
intra- and inter - wban access scheme : time division multiple access(tdma ) across multiple wbans is employed as well as within the intra - wban access scheme since it provides better interference mitigation with respect to power consumption and channel quality @xcite ; 2 .
three - branch cooperative communication : a wban system in this paper uses two relay nodes to provide extra diversity gain at the receiver .
three branches are used with one direct link from sensor to gateway node and two additional links via two relays ; 3 .
intra - wban channel model : extensive on - body channel gain measurements are adopted as the inter - wban channel model for the analysis by simulation .
there are also some major differences in our simulation setup : 1 .
diversity combining scheme : instead of a three - branch selection combining ( sc ) scheme at the gateway node @xcite , an opportunistic relaying ( or ) method is simulated . or reduces complexity when compared with sc by adopting the concept that only a single relay with the best network path towards the destination forwards a packet per hop ; 2 .
inter - wban channel model : a major contribution in this paper is that performance of cooperative wbans using opportunistic relaying with no cooperation between wbans is investigated with real inter - body channel gain measurements . in a multi - wbans co -
existence scenario , a wban s performance is more interference limited .
therefore , the effectiveness of the proposed scheme is analysed based on signal - to - interference - and - noise ratio ( sinr ) .
since the duration of outage is more important than the probability of outage in analyzing the performances of a communication system with multiple co - channel interferers @xcite , then here our analysis shows both the first order statistics of outage probability and second order statistics of level crossing rate ( lcr ) for sinr .
statistics are compared with respect to a traditional star topology wban with the same channel data employed .
the model of a wban system used in the simulation consists of one hub ( gateway node ) , two relay nodes and three sensor nodes , which are organized in a star topology . in order to cooperatively use measured channel data , the hub and two relays
are placed separately at one of three locations : chest , left and right hips .
for example , fig .
[ fig : wban configuration ] shows the situation where hub is at chest and relays are at left and right hips respectively . in terms of sensor devices , they are located at any of the rx places listed in table [ node location ] . within each wban ,
sensor nodes are coordinated by the hub using a time division multiple access ( tdma ) scheme .
therefore , as soon as the hub sends out a beacon signal to all connected sensors , they respond by transmitting required information back to hub in a pre - defined sequence during their allocated time slots . during the transmission period ,
one of the relays assists transmission by providing another copy of the signal to the hub .
the choice of relay is based on an opportunistic relaying scheme that will be explained in a later section . in this paper
, it is assumed that only one information packet is sent from each sensor during its allocated time slot .
after completion of transmission from all sensor nodes , the system becomes idle until the next beacon period . for best co - channel interference mitigation and for reducing power consumption , tdma is adopted as the co - channel access scheme across wbans , as well as being the access scheme within any wban .
assume there are total @xmath0 wbans co - located ( or in close proximity ) and the number is fixed during the period of simulation , then the shared channel can be evenly divided into @xmath0 time slots with a length of @xmath1 .
therefore , each wban goes into idle status for a length of @xmath2 , which is also denoted as @xmath3 , after it completes transmission .
however , due to lack of a global coordinator among multiple wbans , the execution of the tdma scheme used here is slightly different from the traditional implementation of tdma . here , in our tdma scheme
the coordinator in each wban chooses the transmission time of every superframe randomly , following a uniform random distribution over @xmath4 $ ] . in this paper , two wbans employing this same configuration
are used in analysis .
the extensive on - body wban channel data was measured with small wearable channel sounders operating at 2.36 ghz over several hours of normal everyday activity of an adult subject .
this process was repeated several times with devices on different experiment subjects .
the measuring system on a single subject consisted of three transceivers and 7 receivers .
their locations are listed in table [ node location ] . according to the experiment setup
as shown in table [ node location ] , samples were taken over a period of two hours . during this process , three transceivers broadcasted in turn at 0 dbm in a round - robin fashion , with each one occupying the channel for 5 ms .
hence , each transceiver transmits every 15 ms .
while one was transmitting , the remaining channel sounders recorded the received signal strength indictor ( rssi ) if a packet was successfully detected . .tx
/ rx radio locations , x indicates a channel measurement .
lh - left hip , rh - right hip , c - chest , hd - head , rw - right wrist , lw - left wrist , lar - upper left arm , la - left ankle , ra - right ankle , b - back [ cols="^,^,^,^,^,^",options="header " , ] [ simulation combination ] outage probabilities at given sinr thresholds are defined as the probability of sinr value being smaller than a given threshold @xmath5 the performance of single link communication and two - hop cooperative communication schemes are compared with respect to sinr threshold at 1% and 10% outage probability . here
, 10% outage probability corresponds to a guideline for 10% maximum packet error rates in the ieee 802.15.6 ban standard @xcite .
[ fig : outage probability ] shows the outage probability for situations when different subjects are treated as the person - of - interest .
note that in fig .
[ fig : subjectone_outage ] , the relevance of choosing a starting sample index @xmath6 is explained in the next paragraph . for subject one
, the cooperative communication scheme provides about @xmath7 db improvement over traditional single link communications at 10% outage probability . at 1% outage probability ,
the improvement is even more significant , there is , in fact , more than 15 db improvement .
in contrast , simulation on subject two shows similar results , with 4 db and 6 db improvement at 10% and 1% outage probabilities respectively . however , while running simulations using a different starting channel sample index , it is found that channel stability has a significant impact on the performance of the cooperative communication scheme .
[ fig : channel gain plot ] is a typical on - body channel gain plot for subject one .
it is clear that there is a significant change in operating environment at the point where the red line is placed in fig .
[ fig : channel gain plot ] .
subject one s on - body channels become more unstable after that point , i.e. channel coherence time decreases rapidly .
hence , simulation is also performed with channel sample taken from @xmath8 , and result is plotted in fig .
[ fig : outage probability of subject one first half ] . compared to fig .
[ fig : subjectone_outage ] where channel sample is taken from @xmath6 , cooperative communication scheme provides little sinr improvements at both 10% and 1% outage probabilities in this situation . as a result
, two - hop cooperative communication schemes can provide even more improvement of the system performance when the environment in which the wban / s are operating changes rapidly .
level crossing rate(lcr ) of time - varying sinr is defined as the average frequency of a received packet s sinr value going below a given threshold in the positive direction .
assume there are a total of @xmath9 crossings at threshold @xmath10 , then the corresponding lcr value is calculated as @xmath11 where @xmath12 is the time between the @xmath13th and @xmath14th crossing .
[ fig : lcr of subject one second half ] and [ fig : lcr of subject two ] show that cooperative two - hop communications are able to reduce level crossing rate at low sinr threshold values significantly . at lcr of 1 hz
, the sinr threshold value for subject one rises by an average of 6 db as shown in fig .
[ fig : lcr of subject one second half ] with the use of cooperative communications . for subject two , the improvement , as shown in fig .
[ fig : lcr of subject two ] , is about 4 db .
system performance remains similar at high sinr threshold values .
in addition , it can similarly be observed from fig .
[ fig : lcr of subject one first half ] in terms of outage probability , when coherence time of the on - body channels in the wban - of - interest is large ( i.e. , channels are more stable ) . in the case of fig .
[ fig : lcr of subject one first half ] , the lcr curves overlap most of the time for single - hop communications and cooperative communications schemes . therefore , in such a case , there is no real performance advantage in terms of level crossing rate to use our proposed cooperative communication scheme .
in this paper , a three - branch opportunistic relaying scheme was investigated in a wban - of - interest under the circumstance where multiple wbans co - exist non - cooperatively .
tdma was employed , as a suitable inter - network , as well as intra - network , multiple access scheme .
empirical inter - wban and intra - wban channel gain measurements were adopted to enable simulation of a practical working environment of a typical wban .
performance was evaluated based on outage probability and level crossing rate of received packets sinr values . it has been found that opportunistic relaying can provide an average of 5 db improvement at an sinr outage probability of 10% .
it also reduces the level crossing rate significantly at low sinr threshold values .
however , the performance of opportunistic relaying relies on the quality of the on - body channel greatly .
opportunistic relaying is particularly more useful , than single - link star topology communications , in a rapidly changing environment .
d. smith and d. miniutti , `` cooperative body - area - communications : first and second - order statistics with decode - and - forward , '' in _ wireless communications and networking conference ( wcnc ) , 2012 ieee _ , paris , france , apr .
r. derrico , r. rosini , and m. maman , `` a performance evaluation of cooperative schemes for on - body area networks based on measured time - variant channels , '' in _ communications ( icc ) , 2011 ieee international conference on _ , june 2011 , pp .
p. ferrand , m. maman , c. goursaud , j .-
m . gorce , and l. ouvry , `` performance evaluation of direct and cooperative transmissions in body area networks , '' _ annals of telecommunications _ , vol .
66 , pp . 213228 , 2011 .
y. chen _
et al . _ ,
`` cooperative communications in ultra - wideband wireless body area networks : channel modeling and system diversity analysis , '' _ selected areas in communications , ieee journal on _ , vol .
27 , no . 1 ,
5 16 , jan . 2009 .
j. dong and d. smith , `` cooperative body - area - communications : enhancing coexistence without coordination between networks , '' in _ personal , indoor and mobile radio communication ( pimrc ) , 2012 ieee _ , sydney , australia , september .
a. zhang , d. smith , d. miniutti , l. hanlen , d. rodda , and b. gilbert , `` performance of piconet co - existence schemes in wireless body area networks , '' in _ wireless communications and networking conference ( wcnc ) , 2010 ieee _ , sydney , australia , apr .
2010 , pp . 1 6 .
r. annavajjala and j. zhang , `` level crossing rates and average outage durations of sinr with multiple co - channel interferers , '' in _ military communications conference , 2010 - milcom 2010 _ , 31 2010-nov . 3 2010 , pp .
1233 1238 .
d. smith , l. hanlen , and d. miniutti , `` transmit power control for wireless body area networks using novel channel prediction , '' in _ ieee wireless communications and networking conference ( wcnc _ , paris / france , april 2012 . | in this paper , a cooperative two - hop communication scheme , together with opportunistic relaying ( or ) , is applied within a mobile wireless body area network ( wban ) .
its effectiveness in interference mitigation is investigated in a scenario where there are multiple closely - located networks . due to a typical wban s nature
, no coordination is used among different wbans . a suitable time - division - multiple - access ( tdma )
is adopted as both an intra - network and also an inter - network access scheme .
extensive on - body and off - body channel gain measurements are employed to gauge performance , which are overlaid to simulate a realistic wban working environment .
it is found that opportunistic relaying is able to improve the signal - to - interference - and - noise ratio ( sinr ) threshold value at outage probability of 10% by an average of 5 db , and it is also shown that it can reduce level crossing rate ( lcr ) significantly at a low sinr threshold value .
furthermore , this scheme is more efficient when on - body channels fade less slowly . |
gamma - ray bursts ( grbs ) are among the most powerful objects in the universe .
the energy output in the @xmath5-rays is in the order of @xmath6 ergs ( assuming isotropic emission ) on second time - scales at cosmological distances @xcite .
grbs are divided into two classes , short and long , which are usually defined by @xmath7 @xcite , the time during which the cumulative counts in the detector increases from 5@xmath8 to 95@xmath8 above the background rate .
there are two types of plausible central engine model proposed for grbs : ( 1 ) stellar mass black holes , which accrete material from a remnant star with an extremely high accretion rate ( @xmath0 ( 0.1 - 1 ) @xmath9 @xmath10 ) @xcite ; ( 2 ) rapidly rotating ( @xmath0 ms duration ) , highly magnetised neutron stars ( a proto - magnetar ) @xcite . based on the latter model of the central engine
, @xcite proposed that grbs may be accompanied by very bright , short duration ( @xmath11 ms ) radio bursts , which could be observed at frequencies of tens of mhz .
detection of such radio emission from a grb at a cosmological distance could provide a probe for understanding both the ionised intergalactic medium ( igm ) and grb physics @xcite .
however , @xcite showed that observations of short duration radio pulses from grbs at low radio frequencies ( @xmath12mhz ) would be significantly limited by scattering effects because the radio emission passes through the dense stellar wind of the progenitor or the immediate interstellar medium surrounding the source .
@xcite showed that these propagation effects alter the properties of radiation with brightness temperature @xmath13k , which may hinder the detectability of any radio pulse .
recently , @xcite proposed that grbs ( with a magnetar as the central engine ) could produce short duration radio pulses when the supra - massive neutron stars undergo rapid spin down ( due to strong magnetic fields ) within @xmath14 s of their birth , collapsing into black holes as they lose centrifugal support .
these short duration radio pulses are proposed to be detectable at radio frequencies ranging from @xmath15 to @xmath16 ghz , if they are not absorbed by the grb blast wave in front of the emission region .
@xcite suggests that a small fraction of short duration radio pulses could be physically connected with grbs . in recent years
, there have been number of detections of non - repeating short duration radio pulses .
the dispersion measures ( dm ) of these pulses are so large that the progenitors are suggested to be extragalactic in origin .
@xcite reported the discovery of the first such short duration radio pulse ; they discovered a single pulse of duration @xmath17 5ms at a dm of 375 pc @xmath1 from archival data @xcite recorded using the 64 m parkes radio telescope .
@xcite inferred an event rate of 50 @xmath18 ( assuming an isotropic distribution of the sources in the sky ) and proposed that the implied rates of occurrence were compatible with grb rates .
however , there are no recorded grb events satellite was observing at the time and position of the event . ] near the pulse location .
another event was discovered by @xcite from archival data @xcite with the same telescope .
the pulse properties were similar to that reported by @xcite , with a short time scale ( @xmath0 7 ms ) and high dm ( 745 pc @xmath1 ) .
recent analysis of this pulse has suggested it is galactic in origin @xcite some doubts on the extragalactic origin of these pulses have been put forward , based on the work of @xcite , with the detection of 16 similar highly dispersed pulses from parkes archive data @xcite , all deemed as terrestrial in origin .
most recently , however , the evidence for an extragalactic origin has strengthened significantly , with the discovery of four new highly dispersed pulses by @xcite .
@xcite coined the term fast radio burst ( frb ) to describe these bursts .
the frbs were detected with dms in the range 553 to 1103 pc @xmath1 . based on their properties , @xcite infers the frb event rate ( @xmath19 ) to be @xmath20 in the comoving sample volume that contains @xmath21 late type galaxies , which is @xmath0 @xmath22 times larger than the grb rates ( @xmath23 = @xmath24 ) . using the inconsistency in the event rates and , in particular , with no known grb events coincident with the frbs , @xcite rules out grbs as possible progenitors for most frbs .
all six frbs mentioned above were detected by the 64 m parkes radio telescope and its 13-beam receiver , creating concerns about the astrophysical origin of the pulses .
most recently , @xcite reported a discovery of a seventh frb in the 1.4 ghz pulsar alfa survey with the 300 m arecibo telescope , the first frb discovered from a geographical location other than parkes .
this frb was found at low galactic latitude ( @xmath25 ) with a dm of @xmath26 and pulse width @xmath27 . despite the low galactic latitude
, the high dm suggests an extragalactic origin .
none of the frbs from parkes or arecibo to - date have been associated with a grb , although the theoretical possibility that they might be connected has motivated several experiments to look for coincident grb / frb emission .
* hereafter b12 ) performed a targeted survey to detect short duration radio pulses from grbs at 1.4ghz .
b12 observed nine grbs , both long and short , using a 12 m radio telescope located at the parkes observatory .
b12 reported the detection of two single radio pulses from two long grbs with s / n @xmath28 6@xmath29 .
the pulses were detected 524 and 1076 seconds after the @xmath5-ray emission from the grbs .
the dms of the pulses were 195 pc @xmath1 and 570 pc @xmath1 , which are large compared to the expected galactic electron - density contribution @xcite to the dm , indicating the possibility that the bursts are extragalactic in origin and possibly from the grbs .
similar to the b12 experiment , there have been a number of past and ongoing experiments searching for frb - like emission from the prompt phase of grbs at low and high radio frequencies @xcite .
@xcite observed 19 grb events at 151mhz between 1970 and 1973 , with a time resolution of 0.3s ; they searched for coincident radio pulses with two or more stations over periods @xmath30 to @xmath31 relative to the grb event times .
their search failed to locate any astronomical radio pulse above their sensitivity limit of @xmath32 .
two other grbs were observed by @xcite , who performed an automatic rapid radio follow - up experiment at 151mhz , with a time resolution of 1.5s ; they reported a non - detection at a sensitivity of @xmath33 of 73jy .
@xcite observed 7 grbs between 1992 and 1994 , covering a large range of frequencies between 40 and 1000mhz using three similar spectrometers , with a time resolution of 0.15s .
although the experiment sensitivity was low ( @xmath32 ) , the spectral coverage made this experiment very interesting .
similarly , @xcite observed 32 grbs at 73.75mhz from 1997 to 1998 and had a wide range of sensitivity for each grb .
they reported a non detection at sensitivity limit of @xmath34 for integration times of @xmath35 .
most recently , @xcite reported the observations of 12 grbs at 74mhz , five grbs at 52mhz , and 17 grbs at 37.9mhz with a time resolution of 5s .
similarly , @xcite reported the observation of four grbs at 15ghz with a time resolution of 0.5s .
both @xcite and @xcite did not detect any frb - like emission from the observed grbs .
the sensitivities of the above mentioned experiments were generally low , however they have motivated a number of follow - up experiments including this work and a recent follow up project proposed by bannister et al ( 2013 ) ( vlba project codes bb317 and bb325 ) to use one or more very long baseline array ( vlba ; @xcite ) dishes to automatically follow up grb alerts during telescope idle time .
the vlba - grb experiment searches for frbs using the existing v - fastr system @xcite . in this paper , we describe observations performed at 2.3ghz using the hobart 26 m radio telescope to search for frbs associated with grbs .
we designed our experiment to be similar to the b12 experiment with some improvements .
our aim was to observationally test the possible association of frbs with grbs and compare our results with the b12 results .
section [ sec : observations ] of this paper describes the observations performed .
a detailed description of our single pulse search pipeline and verification is given in section [ sec : dataprocessing ] .
in section [ sec : resultsanddiscussion ] we analyse our results and compare them with the b12 results .
conclusions are presented in section [ sec : conclusion ] .
lllcllllllc source & @xmath36 & @xmath37 & @xmath38 & trigger & @xmath39 & @xmath40 & @xmath7 & @xmath41 & @xmath42 & @xmath43 + & ( j2000.0 ) & ( j2000.0 ) & ( @xmath44 ) & date & ( ut ) & ( ut ) & ( s ) & ( ut ) & ( s ) & ( ut + s ) + grb 111212a & @xmath45 @xmath46 @xmath47 & @xmath48 @xmath49 @xmath50 & 3 & 2011-dec-12 & 09:23:07 & 09:24:06 & 30.86 & 09:24:50 & 1200 & @xmath39 + 2856 + grb 120211a & @xmath51 @xmath52 @xmath53 & @xmath54 @xmath55 @xmath56 & 3 & 2012-feb-11 & 11:58:28 & 11:59:29 & 101 & 12:03:10 & 1800 & @xmath39 + 119.5 + grb 120212a & @xmath57 @xmath58 @xmath59 & @xmath60 @xmath61 @xmath62 & 3 & 2012-feb-11 & 09:11:22.74 & 09:11:43 & 20.42 & 09:13:20 & 1800 & @xmath39 + 2676 + grb @xmath63 & @xmath64 @xmath65 @xmath66 & @xmath67 @xmath68 @xmath69 & 3 & 2012-feb-18 & 00:49:22.14 & 00:50:17 & 7 & 00:52:10 & 1800 & - + grb 120224a & @xmath57 @xmath70 @xmath71 & @xmath72 @xmath73 @xmath56 & 2.4 & 2012-feb-24 & 04:39:56.49 & 04:40:40 & 2.06 & 04:42:00 & 1800 & @xmath39 + 109.1 + we observed the five grbs listed in table [ tbl-1 ] at a central frequency of 2.276ghz , using the 26 m radio telescope located at the mount pleasant radio observatory operated by the university of tasmania near hobart , tasmania .
a computer at the observatory is configured to receive grb notifications via email from the gamma - ray coordination network ( gcn ) .
the gcn system distributes the location of the grbs detected by space - borne observatories such as integral @xcite , swift @xcite , agile @xcite and fermi @xcite in real time .
a filter is applied to accept notices only from the swift satellite , hence the grbs listed in the table [ tbl-1 ] are all observed by the swift satellite . the gamma - ray burst alert telescope ( bat ) on swift
first detects the grb and localizes the burst direction to an accuracy of @xmath74 .
these are the co - ordinates used by the telescope control system . given that the radio telescope s primary beam full width at half maximum ( fwhm ) for this experiment is about @xmath75 , all the observed grbs in this experiment fall within the beam . only grbs that are above the elevation limit ( @xmath76 ) and within the visible range of the telescope ( @xmath77 @xmath78 @xmath37 @xmath78 @xmath79 )
are observed .
when an accessible grb is identified , the telescope slews to the listed coordinates and then starts recording data . on average
it takes approximately 140 seconds for the telescope to slew to the required location .
data were recorded using the australian long baseline array ( lba ) data acquisition system ( das ) @xcite .
the lba das has a sampler and digital filter which receives two independent analog intermediate frequencies ( ifs ) , the right and left circular polarised signals over a 64mhz bandwidth .
the analog ifs are sampled and digitised to 8 bit precision and then re - sampled to 2 bits using digital filters @xcite before the digital representation of the voltage is recorded to computer disk . for each observation the system equivalent flux density ( sefd ) of the telescope was measured to be 800 - 900jy , which is calculated as @xmath80 where @xmath81 is the system temperature ( 92 @xmath82 103k ) , @xmath83 is the boltzmann constant , @xmath84 and @xmath85 are the antenna efficiency ( 0.6 ) and the diameter of a circular aperture antenna ( 26 m ) , respectively .
the automatic gain control loops were turned off for the duration of the observations .
the one sigma flux density sensitivity for the 26 m telescope ( to detect a pulse of width 25ms ) is approximately 0.5jy , calculated as @xmath86 where @xmath87 and @xmath88 are the bandwidth ( 64mhz ) and integration time ( 25ms ) , respectively , and @xmath89 is the number of polarisations ( 2 ) .
the pointing error of the 26 m telescope across the whole sky is @xmath90 .
the observations occurred over a period of 30 minutes for each grb except for grb 111212a which was observed for 20 minutes .
following observation of the grb field , the telescope slewed to a nearby calibrator and data were recorded for 20 minutes . a control observation
was then undertaken by pointing the telescope at blank sky ( @xmath91 away from the grb position ) for approximately 20 minutes and recording the voltage data .
s to 25.60ms ( section 3.1.2 ) .
each averaged time series go through the rfi excision , dedispersion , event detection and clasification stages independently .
[ flowchart],width=576,height=576 ]
the recorded data were transferred to the petabyte data storage facility at ivec , in perth .
data processing was performed on the curtin university parallel processor for astronomy ( cuppa ) , a 20 node computer cluster operated by curtin university . in this section
we describe the data processing pipeline ( figure [ flowchart ] ) used to search for single short duration radio pulses in our data .
the search pipeline has multiple stages : correlation ; time averaging ; radio frequency interference ( rfi ) excision ; dedispersion ; event detection ; and classification of events .
several of our tools were reused from the v - fastr experiment @xcite which can read lba data natively .
the first step in the processing pipeline was to pass the data through a spectrometer to form channelised total power for each polarisation .
the 2-bit digitised voltage samples from each polarisation were decoded and sent to a software correlator .
the correlator is an fx design with a configurable number of spectral channels and averaging time .
we used @xmath92 mhz spectral channels , formed via fft , and an averaging time of 640@xmath2s .
the output of the software correlator was written to files for subsequent processing .
the intrinsic width of pulses are generally unknown . because the dispersion and scattering affects the observed width of the pulse , a large parameter space of pulse widths must be searched @xcite . to obtain optimal sensitivity ,
the correlated time series were averaged for pulse widths ranging from 1.28ms to 20.48ms .
this was achieved by averaging @xmath93 time samples together , where @xmath94 ranges from 1 to 5 , and searching each averaged time series independently .
an additional time averaged series was generated at 25.60ms by averaging 40 time samples together ( which is not a power of 2 ) , to match the width used by b12 in their experiment .
observations at 2.3ghz are generally affected by radio frequency interference ( rfi ) from terrestrial transmitters , radars , satellites etc . , which changes the total power as a function of frequency and time .
therefore , identification and excision of rfi was undertaken .
first , we excise the time samples in each frequency channel that were affected by sporadic rfi .
the top panel of figure [ rfi ] shows an example of sporadic , wide - band rfi , where the power level in all the frequency channels over short time scales ( in this case @xmath0 4ms ) is much higher when compared to the typical mean power in neighbouring time samples across the frequency band .
the top panel of figure [ impulsiverfi ] shows an example of sporadic narrow - band rfi , where the power level in a frequency channel over a short time ( in this case @xmath95s ) is high when compared to the typical mean power in the neighbouring time samples in that frequency channel .
our rfi excision algorithm used a thresholding method to identify samples in the time domain which were affected by rfi .
first , we calculated the local variance ( @xmath96 ) of the time samples in each frequency channel over an adjustable time range ( @xmath0 10 sec ) and then compared to the power ( amplitude ) of each time sample .
the time samples that exceeded a pre - set threshold ( here the threshold was set to @xmath97 ) were excised from the time series and replaced with the mean power value estimated for each frequency channel over the band .
we conservatively chose a 5@xmath98 threshold for rfi excision , because we did not want to clip the thermal noise distribution is @xmath99 and for @xmath100 the probability is @xmath101 .
we have @xmath102 independent samples , therefore , at @xmath103 we would be effectively clipping @xmath0 1 sample as a false positive , compared to when a 3@xmath29 threshold is chosen we would be effectively clipping @xmath02600 samples .
this could skew the thermal noise distribution in the data . ] .
the bottom panels of figures [ rfi ] and [ impulsiverfi ] show the results of our excision algorithm .
+ + next , we manually identified and excised persistent narrow - band rfi which are constant in time over a frequency channel .
the top panel of figure [ rfi ] shows an example of persistent narrow band rfi where the power in a frequency channel is higher compared to the power in the neighbouring channels across the frequency band .
the affected channels , including the lower band edge spectral channel ( dc channel generally has spurious high power due to sampler offsets ) , were recorded in a flagging file .
the dedisperser ( the next stage in the pipeline ) reads the flagging files for bad channels and ignored the bad channels in further processing .
we repeated the rfi excision process on all the time averaged data series .
, title="fig:",width=590,height=201 ] + , title="fig:",width=590,height=201 ] the next step in the pipeline is dedispersion .
any incoming astronomical impulsive radio signal will , in general , be dispersed in time across the spectral channels due to the ionised component of the interstellar and intergalactic media @xcite .
the intrinsic signal can largely be recovered by incoherent dedispersion , by applying an appropriate time delay for each spectral channel and adding power across all the spectral channels .
the time delay between two frequencies is given by @xmath104 , where the dm ( pc @xmath1 ) is the integrated column density of free electrons along the line of sight and @xmath105 and @xmath106 are the low and high frequencies in ghz , respectively @xcite .
we use dart ( dedisperser of autocorrelations for radio transients ) to perform the incoherent dedispersion , a detailed description of which is given by @xcite .
the incoming data to dart are ordered in frequency and time for each polarisation .
the dedispersion process sums the power across channels , with the delay for each channel according to the frequency and dm . at the output
, dart combines both polarisations and generates a dedispersed time series for each dm .
we dedisperse the data with dm steps as listed in table [ dmtable ] ranging from 0 - 5000 pc @xmath1 .
the choice of dm steps was made by setting the time delay ( @xmath107 ) between highest ( @xmath106 ) and lowest ( @xmath105 ) spectral channels equal to the data sampling time ( integration time ) .
choosing dm steps smaller than the data sampling time leads to highly correlated time samples , since the dedispersed time series are virtually identical for neighbouring dm steps .
dart is flexible when compared to the taylor tree algorithm @xcite , because it can dedisperse the signals beyond the diagonal dm , the dm where the gradient of the dispersion curve is one time sample per spectral channel .
the dm steps across the dm range are linearly distributed for the first few dm steps , before progressing geometrically , with a growth rate of 1.02 .
this results in equal spacings at lower dm and increasingly larger spacings at higher dm .
table [ dmtable ] summarises the number of dm steps used for each time average .
ll time averaging ( ms ) & @xmath108 + 0.64 & 149 + 1.28 & 114 + 2.56 & 79 + 5.12 & 45 + 10.24 & 23 + 20.48 & 12 + 25.60 & 9 + the next step in the pipeline is detection .
after dedispersion , each time series is searched for events above a defined signal to noise threshold . in the detector ,
the time series are divided into small time segments ( @xmath109 s ) .
we chose this length of time because visual inspection of the data showed that the total power sometimes changed on time - scales longer than this .
if longer time segments were used this would artificially raise the variance estimate .
each time segment is high - pass filtered , by subtracting a 100 timestep moving average from each sample .
this reduces any long term power variations due to system temperature ( @xmath81 ) changes - atmosphere , receiver response etc . , which may bias variance estimates on time scales of 100s of samples , while retaining sensitivity to short time scale variations .
we note that the high - pass filter window is longer than the expected pulse width of an astronomical event .
each dedispersed time segment is then searched for single events @xmath28 5@xmath29 .
the noise level of the time series for each dm step is estimated using the standard deviation .
we also tried the robust trimmed estimator , interquartile range ( iqr ) @xcite , to estimate the noise level for each dm step .
the number of events detected @xmath110 from the iqr and standard deviation techniques did not differ .
we adopt use the standard deviation because the behaviour of this statistic is well understood when a finite number of samples are used for its estimation . the next step in the pipeline
is the classification of detected events @xmath111 .
figure [ dmtimerfi ] demonstrates the differences between the signatures of astronomical signals , low level rfi and thermal noise fluctuations in the dm / time plane .
an astronomical signal is expected to show reduced s / n for dm steps away from the true dm .
the astronomical signals are broad band and any signals well above the minimum detection threshold are expected to be seen in multiple dm steps for a given time average and also in the consecutive time averages .
events due to thermal noise fluctuations will be seen at a single dm step and follow gaussian statistics ( most of the events would be close to the s / n threshold ) .
events due to low - level rfi will appear in multiple dm steps , but do not have the characteristic peaked shape that a bright pulse from an astronomical source would have .
the recorded events @xmath110 from our detection pipeline for all the time averages were visually examined for signatures of astronomical signals , low level rfi and thermal noise fluctuations using the diagnostic plots and were classified respectively . to test our single pulse search pipeline we used archival observations of the pulsar psr j0835 - 4510 ( the vela pulsar ) .
observations were made using the same 26 m radio telescope as used for our grb observations .
the data recording system used was also identical to the recorder used for the grb observations .
vela was observed at a central frequency of 1440 mhz with right and left circular polarisations over a 64 mhz bandwidth .
vela has a steep spectral index , @xmath112 ( s @xmath113 ) , dm = 67.99 pc @xmath1 and a period of 89.3 ms @xcite .
the recorded voltage data were correlated at 640 @xmath114s time resolution and 128 spectral channels , the same as for the grb data .
the correlated data were dedispersed over a dm range of 0 - 200 ( pc @xmath1 ) , with 40 dm steps .
the dedispersor output was searched for pulses above the s / n threshold of 5 at 640 @xmath114s time averaging and resulting events were examined to determine if they were real pulses from the pulsar in the time / dm plane ( figure [ dmtime ] ) .
individual pulses were detected at dm @xmath0 67 pc @xmath1 and at neighbouring dm steps with reduced s / n as expected .
segments of data were written out for each candidate pulse for further investigation . in the follow - up process , candidates were visually inspected via dynamic spectrum plots similar to those shown in figure [ dynsplot ] .
the dynamic spectrum shows individual dispersed pulses with a period of 89.3 ms in time across the frequency band .
therefore , we have confirmed detection of individual pulses with vela s period and dm via our single pulse search pipeline .
5@xmath29 vs dm ( pc @xmath1 ) as a function of time .
the size of the circles are proportional to the signal to noise ratio .
single pulses from vela are detected at the true dm ( horizontal line ) and across multiple dm steps with reduced s / n .
[ dmtime],width=456,height=336 ] 89.3 ms which is the correct period for vela .
[ dynsplot],width=456,height=321 ]
here we discuss the results of our search for frbs associated with grbs using the single pulse search pipeline .
the five grbs have @xmath115 s based on the light curves from bat .
processing the five data sets and visually inspecting the events via diagnostic plots did not show any signatures of astronomical signals .
all the events detected by our detector were between @xmath4 and @xmath116 and no events @xmath3 were detected .
table[tbl-2 ] lists the number of @xmath117 events for all the grbs and time averages , respectively .
figure [ diag ] shows an example diagnostic plot which illustrates the distribution of detected events @xmath118 for grb 120211a at 1.28 ms time average . in the absence of a characteristic signature of a astronomical signal @xmath119
, we performed a statistical analysis on events between @xmath120 to check if they were consistent with expectations due to thermal noise fluctuations .
this analysis gives us confidence that our search pipeline is working correctly and highlights the importance of sample variance in determining the significance of an event .
for grb 120211a at 1.28 ms time average .
_ top panel _ :
all events with s / n @xmath118 plotted vs dm and time .
the size of circles are proportional to the s / n of the events . _ lower left panel _ : scatter plot of the dm and s / n . _
lower middle panel _ : distribution of the detected events @xmath111 .
_ lower right panel _ :
number of detected events @xmath111 vs dm .
[ diag],width=552,height=321 ] here we compare the number of events expected due to thermal noise fluctuations to the number of events recorded by our pipeline .
the number of events expected due to thermal noise fluctuations , @xmath122 , above the detection threshold is given by , @xmath123 where @xmath108 is the number of independent dms , @xmath124 is the number of independent time samples , and @xmath125 is the probability of an event above 5@xmath29 due to thermal noise fluctuations ( assuming the power fluctuation due to thermal noise follows the gaussian distribution ) , and is given by the cumulative distribution function ( cdf ) , @xmath126 , \label{prob}\ ] ] where erf is the error function and @xmath127 is the gaussian distribution with mean @xmath114 and variance @xmath128 . in equation 1 , we note that the @xmath108 steps should be independent . given our choice of step progression
we expect @xmath0 50@xmath8 of power samples to be identical between consecutive pairs at lower dm steps ( figure [ corr ] ) . here
we perform an analytical calculation showing that the time series in each dm step at the lower end of the dm range can be treated as independent .
our calculation assumes that a 5@xmath29 noise event is detected due to the addition of power sample fluctuations in the time / frequency plane .
based on this assumption we consider the following scenarios : @xmath129 the probability of two consecutive dm steps each having a @xmath103 event , since @xmath0 50@xmath8 of power samples are identical at the lower dm ; and @xmath130 the probability of two dm steps separated by one dm step each having a @xmath103 event , since @xmath0 25 @xmath8 of power samples are identical at lower dm . in the former case ,
the probability is reduced by a factor of @xmath0 100 compared with a single 5@xmath29 event in a dm step due to thermal noise fluctuations . in the latter case
, the probability indicates that events @xmath110 are consistent with events due to thermal noise fluctuations reduced by @xmath131 ( a detailed description of the calculation is given in appendix a ) .
the small probability in the former and latter scenarios implies that @xmath103 detections in two or more dm steps at the lower end of the dm range are independent .
the expected number of events due to thermal noise fluctuations , @xmath132 , is uncertain due to two effects : @xmath133 events detected due to thermal fluctuations are stochastic and their number is characterised by the poisson distribution ; @xmath134 the variance measured from the data is a sample estimate , which has inherent uncertainty on the variance as it is being estimated over a finite number of samples .
we take these effects into account in order to make our statistical analysis robust .
@xmath133 _ stochasticity of 5@xmath29 events _ : we note that the thermal noise is stochastic in nature , inducing an uncertainty in the estimation of @xmath132 which can be quantified using the poisson distribution .
the probability of observing x events , when @xmath135 number of events is expected , is given by @xmath136 for large @xmath135 , the poisson distribution approximates the gaussian distribution where @xmath137 @xcite .
therefore , for @xmath138 we use the gaussian approximation and the uncertainty in the number of events fluctuates around the mean ( @xmath139 ) with a standard deviation @xmath140 . for @xmath141 , the gaussian approximation is invalid , and we use monte carlo simulations to estimate the uncertainties .
this is described in section 4.2.2 @xmath134 _ sample variance in estimating true signal variance _ : a finite number of samples ( n = 2048 time samples at the intrinsic temporal resolution ) is used to estimate the mean and the standard deviation of the data . the true mean and standard deviation of the data
are not known .
this estimate has an associated uncertainty known as sample variance , because only a finite number of samples is used .
the standard error on the estimation of the mean ( @xmath142 ) and variance ( @xmath143 ) for n independent samples from the true value is given by @xmath144 and @xmath145 respectively .
the errors @xmath142 and @xmath143 are propagated into the expression for the cumulative probability ( equation 2 ) and the variance on the cumulative probability is obtained .
for a detection threshold of @xmath103 , the variance is given by @xmath146 , @xmath147 where n is the number of time samples .
the sample size must be large enough to minimize the sample variance , but should not be so large that the mean changes with in the sample which would artificially increase the variance estimate .
the uncertainty in the expected number of events due to sample variance is propagated to an uncertainty in @xmath132 using equations 1 and 4 , @xmath148 the two uncertainties in the estimate of @xmath132 are independent and gaussian distributed ( section 4.2.1 ) , and therefore can be combined in quadrature , @xmath149 in the previous section , we combined the two uncertainties associated with the estimation of @xmath132 in quadrature , since both uncertainties obey gaussian approximations . here
we would like to note again that the gaussian approximation in case 1 is valid only for @xmath151 .
we observe from the table 3 and 4 that for higher time averages ranging @xmath152s , @xmath141 , the gaussian approximation for the uncertainty discussed in the section [ subsec : uncertainties ] is no longer valid .
therefore , for higher time averages ( @xmath152s ) , we can not combine both uncertainties in quadrature . to estimate the significance @xmath153 of the observed events ( @xmath150 )
, we split the calculations into two regimes : @xmath154 at lower time averages ranging @xmath155s - 2.56ms were @xmath151 , we calculate @xmath153 via the gaussian approximation with mean @xmath132 and variance @xmath156 , given by the equations [ eq : one ] and [ eq : deltae ] respectively ; @xmath157 at higher time averages ( @xmath152s ) where @xmath141 , we use results obtained via monte carlo simulation to estimate @xmath153 . in both regimes , we interpret a @xmath158 to be consistent with the null hypothesis which is that the observed events @xmath150 are consistent with noise . in the low time average regime @xmath153 is given by , @xmath159 to determine the significance of the observed events @xmath160 at higher time averages ( when @xmath161 ) , we use the monte carlo ( mc ) technique and perform @xmath162 simulations of random data sets .
the simulations were performed on the time averages in the range @xmath163 .
the results of the simulations automatically include the combined uncertainty due the thermal noise and sample variance .
our program generates an array of gaussian distributed random data sets of duration 30 minutes and @xmath164 dm steps ( where @xmath164 is given in table [ dmtable ] ) . the array is then passed through the event detection pipeline ( described in detail in section [ sec : dataprocessing ] ) . for each iteration
the mc simulation detects events @xmath111 .
we performed @xmath165 iteration for each time average .
figure [ hist ] shows an example histogram for 5.12 and 10.28 ms time averages where we plot the events detected over @xmath165 iterations .
the values in each bin are normalised to plot the probability density function , such that the integral over the range is unity .
the significance ( @xmath153 ) , is determined by integrating the values in the histogram greater than and equal to @xmath150 value for a given time average .
for instance from table [ tbl-2 ] , the number of events , @xmath150 , recorded for grb120211a is three at 5.12 ms time resolution .
the significance of @xmath150 at the 5.12 ms time resolution can be calculated via the histogram summing the values greater than and equal to the bin number three ( solid green line figure [ hist]a ) .
the region in the histogram greater than and equal to @xmath150 value is termed the critical region in statistics . integrating the point probabilities greater than and equal to the critical value gives the significance @xmath153 @xcite .
@xmath132 for each higher time average is given by the mean of the distribution , which is calculated by taking the average of values obtained through mc simulations ( dotted lines in figure [ hist ] ) . here
the duration of the random data sets was chosen to be the same as the observation duration for each grb and blank sky data set .
the statistical analysis was applied to all the grbs for all averaging times .
table [ tbl-2 ] presents the results of the analysis . in all cases
@xmath166 , indicating that our recorded events are consistent with thermal noise fluctuations . in the absence of a characteristic signature of an astronomical pulse @xmath119
, we conclude that the events @xmath111 are consistent with the expectations due to thermal noise fluctuations .
there were no events with a statistically significant s / n . another approach to determine the probability of an event occurring due to thermal noise fluctuations is to measure the false positive rate when there is no grb in the beam . observing a patch of blank sky after the grb observation preserves the noise properties and rfi information , assuming that no source of astronomical impulsive radio emission is in the field of view .
we analysed the data from the blank sky observations , which were processed in a manner identical to that described for the grb data .
table [ tbl-3 ] shows the results of our analysis for the blank sky datasets .
again , in all the cases @xmath166 indicating that our recorded events are consistent with thermal noise fluctuations . in this section
we compare our results with the b12 and other previous experiment results , to gain some understanding of the limits on frb associated with grbs and to give a quantitative assessment of the conclusions in b12 .
we have designed our experiment to be similar to b12 and have also made significant improvements .
firstly , unlike b12 we have undertaken control observations using blank patches of sky @xmath0 2 degrees away from the grb positions .
secondly , the statistical analysis of the events detected from the blank sky gives estimates for the number of false positives due to thermal noise fluctuations and rfi .
this method is more reliable when compared to the method adopted by b12 to account for the false positive rate by randomising the spectral channels , using data from the grb observations .
b12 showed that randomising the spectral channels would preserve the properties of thermal noise fluctuations , narrow band and wide band rfi , but the presence of impulsive rfi that spans multiple channels are destroyed .
this will reduce the number of false positive events due to thermal noise fluctuations and rfi .
this effect will be high if the rate of impulsive rfi that spans over multiple channels is high , which leads to an inaccurate estimation of false positives .
however , the data from blank sky taken @xmath0 2 degrees away from the grb positions preserves all the properties of thermal noise fluctuations and rfi including the periodic and aperiodic rfi that spans multiple channels .
also , the data go through the same rfi excision and detection algorithms as those used for the grb data sets , which should provide a more reliable estimate of the false positive rates .
the average response times for our experiments and b12 are comparable .
table [ tbl-5 ] compares the parameters of the b12 observations and our observations .
we calculate that the one sigma flux density for our observations to detect a pulse width of 25 ms is approximately 0.5 jy and for the b12 observations is 1.07 jy , indicating that our observations are twice as sensitive as those by b12 , ( although at different frequencies ) .
the two pulses detected by b12 are of @xmath167 millisecond duration .
therefore , the difference in the minimum time resolution in both experiments is not likely to be significant .
.*observational parameters of this work and b12 [ tbl-5 ] * [ cols="<,^,^",options="header " , ] figure [ plightcurve ] shows the x - ray light curves for the grbs observed in our experiment and figure 5 in b12 shows the x - ray light curves for the two grbs for which radio pulses were detected by b12 . the x - ray flux ( @xmath168 , erg @xmath169 ) is given as @xmath170 where @xmath171 is the spectral index , @xmath172 is the temporal decay index , t and @xmath173 are the time and frequency respectively . evaluating figure [ plightcurve ] and figure 5 in b12 , the x - ray telescope ( xrt ) typically starts observations @xmath174 after the prompt @xmath5-rays .
the x - ray flux follows a typical pattern which comprises three distinct power - law segments as described by @xcite , @xcite , and @xcite : @xmath133 initial steep decay slope ( @xmath175 with @xmath176 and t @xmath177 s ) ; @xmath134 a very shallow decay ( @xmath178 and t @xmath179 s ) ; and @xmath180 a steep decay ( @xmath181 and t @xmath182 s ) .
these segments are separated by two corresponding break times , @xmath183 500 s and @xmath184 s. table [ tbl-6 ] lists the temporal indices ( @xmath185 , @xmath186 , @xmath187 ) and spectral index ( @xmath188 ) parameters for all the x - ray light curves in figure [ plightcurve ] and figure 5 in b12 .
the temporal indices and spectral indices for all the grbs listed in the table [ tbl-6 ] are consistent with the standard grb x - ray light curve , indicating that the grbs are from similar populations ( although in many cases we do not see all three power - law segments , due to limited xrt coverage ) .
therefore we note that the grbs from which b12 claim frb emission do not appear different from the grbs we observed in our experiment .
lcccccc source & @xmath185 & @xmath189 ( s ) & @xmath186 & @xmath190 ( s ) & @xmath187 & @xmath191 + grb @xmath192 & 3.09 @xmath193 0.58 & 502 & 0.38 @xmath193 0.30 & 1157 & 0.90 @xmath193 0.08 & 1.12 @xmath193 0.10 + grb @xmath194 & 2.54 @xmath193 0.70 & 708 & - & - & 0.89 @xmath193 0.17 & 0.81 @xmath193 0.30 + grb @xmath195 & - & - & - & - & 1.49 @xmath193 0.12 & 1.03 @xmath193 0.14 + grb 120211a & 1.74 @xmath193 0.18 & 521 & -0.19 @xmath193 0.25 & 5082 & 0.79 @xmath193 0.33 & 0.22 @xmath193 0.21 + grb @xmath196 & - & - & -&- & 1.01 @xmath193 0.07 & 1.08 @xmath193 0.16 + grb @xmath197 & -&-&-&- & - & - + grb 120224a & 7.38 & 136 & 0.01 @xmath193 0.10 & 5348 & 1.07 @xmath193 0.26 & 1.22 @xmath193 0.14 + ) and end time are marked as vertical green lines .
the black lines are the xrt powerlaw fits .
wtslew / wt - settling is the window timing mode during the spacecraft slew .
pc is the photon counting mode . for grb 120218a no xrt lightcurve due to sun observing constraints . for grb 111212a and grb 120218a
no xrt prompt phase information was recorded due to swift satellite orbital constrains .
as the swift satellite is in a low earth orbit with the orbit period of 94 minutes , it suffers from 50@xmath8 time off target per orbit .
[ plightcurve],width=384,height=456 ] , spectral index . _ bottom panel _ : the probability b12 detecting one real event and zero events in other experiments as a function of @xmath198 , spectral index .
@xmath199 is the low flux density cutoff .
the three curves show the limit imposed by adding more experiments at different frequencies .
the red solid line depicts the total probability for p14 and b12 experiments for the range of spectral indices .
the blue solid line depicts the total probability of b12 finding events while d96 , o14 and p14 experiments have found no events .
the green solid line depicts the total probability of finding events while s13 , d96 , o14 and p14 experiments have found no events.[fig10],width=456,height=288 ] here we estimate the probability that the combined experiments detected frb emission , over the combined observations reported in this experiment ( hereafter , p14 ) and in b12 .
since many astrophysical objects follow a luminosity function with power - law exponent @xcite , we assume grbs are distributed homogeneously in a static euclidean universe and each grb generates an frb with differential number density per unit flux given by , @xmath200 where @xmath201 . in order to compare events at different frequencies ,
we assume the flux density of the frb has a spectral index @xmath202 over radio frequencies .
based on these assumptions , we calculate the total probability of our experiment detecting zero events , and b12 detecting two events out of eight .
the total probability , @xmath203 , for both experiments is given by the joint probability distribution , @xmath204 where @xmath205 is the flux density and @xmath206 and @xmath207 are the threshold flux densities for the b12 experiment and our experiment , respectively .
@xmath208 corresponds to the probability of detecting events @xmath209 and is given by , @xmath210 where the denominator normalises the distribution for a lower limit on the frb flux density .
we chose @xmath211 to maximise @xmath203 for @xmath212 , hence to maximize the chance that the observed frbs by b12 are astrophysical .
the choice of @xmath213 corresponds to a low flux density cutoff in @xmath214 that must exist for the probability of observing an frb associated with any grb to not be vanishingly small .
physically this would correspond to a correlation between observed grb flux density and frb flux density . testing
the existence of such a correlation is an underlying motivation for this ( and similar ) experiments .
we note that the observing frequency and s / n threshold are different between our experiment and the b12 experiment ( table [ tbl-5 ] ) .
the intrinsic spectral index is unknown .
this difference is accounted for by scaling the flux density with a range of spectral indices @xmath215 , given by , @xmath216 where @xmath217 is the minimum detectable flux density at 25 ms in b12 ( table [ tbl-5 ] ) , @xmath218 and @xmath219 are the central frequencies in our experiment and b12 , respectively .
next , we estimate @xmath203 in two regimes @xmath133 the probability of detecting zero events in our experiment , and b12 detecting two events ( from eight ) ; @xmath220 the probability of detecting zero events in our experiment , and b12 detecting one event ( from eight ) . in the second case
we calculate the probability using the equation [ eq : prob ] just by changing the upper limit on first two terms of equation [ eq : prob ] . in the former case , the probability @xmath78 0.001 ( top panel of figure [ fig10 ] , solid red line ) and in the latter case the probability @xmath203 @xmath221 0.01 ( bottom panel of figure [ fig10 ] , solid red line ) .
this indicates that non - detections in our experiment are highly unlikely to be consistent with b12 having obtained two detections .
next , we now extend the above analysis to include the experiments of ( * ? ? ? * hereafter d96 ) , ( * ? ? ?
* hereafter s13 ) and ( * ? ? ?
* hereafter o14 ) .
table [ otherexp ] list the parameters for these experiments .
the total probability of observed results over all five experiments , is given by , @xmath222 similar to the above analysis , we calculate the total probability in two regimes : @xmath223 the probability of detecting zero events in four different experiments , and b12 detecting two events ( from eight ) ; @xmath220 the probability of detecting zero events in four different experiments , and b12 detecting one event ( from eight ) . to illustrate the effects of adding the low frequency experiments and the high frequency experiments , we combine the experiments in two steps .
first , low frequency experiments d96 and o14 are combined with b12 and p14 experiments .
next , both high and low frequency experiments s13 , d96 and o14 are combined with b12 and p14 experiments .
the blue solid line in the figure [ fig10 ] depicts the total probability of b12 finding events while d96 , o14 and p14 experiments have found no events .
the green solid line in the figure [ fig10 ] depicts the total probability of b12 finding events while s13 , d96 , o14 and p14 experiment found no events . in both cases
the probability of null detection in s13 , d96 , o14 and p14 experiments , while b12 detecting at least one , is @xmath224 .
the sensitivities of the s13 , d96 , o14 experiments are scaled to detect a 25ms pulse ( table [ otherexp ] ) . in the figure [ fig10 ] , we note a sharp cut - off in the curve , because at this point the low frequency and high frequency experiments have @xmath225 values less than @xmath226 , which implies that the probability of detecting an event would be vanishingly small .
hence , the low frequency experiments ( d96 and o14 ) would have easily detected a 25ms pulse with spectral index @xmath227 given their detection thresholds , and the high frequency experiment s13 would have easily detected a 25ms pulse with the spectral index @xmath228 . finally , we to note that including the other previous low frequency experiments , of @xcite , @xcite and @xcite in the analysis makes no difference to the results .
l l l l c parameters & d96 & @xmath229 & @xmath230 + central frequency ( mhz ) & 151 & 15270 & 74 , 52 , 37.9 + bandwidth & 700 khz & 6 ghz & 75 khz + time resolution ( s ) & 1.5 & 0.5 & 5 + min detectable flux density-1@xmath29 ( jy ) @ 25ms & 20 & 0.026 & 353 , 336 , 311 + @xmath231 of grbs observed & 2 & 4 & 12 , 5 , 17 + s / n threshold & 3 & 3 & 5 +
this experiment searched for frb - like emission from grbs at 2.3 ghz .
we observed five grbs using a 26 m radio telescope , automated to quickly respond to gcn notifications and slew to the source position within minutes .
we did not detect any sigma events @xmath232 similar to b12 , which motivated this experiment .
our analysis of events detected at @xmath28 5@xmath29 in our experiment shows that they are consistent with thermal noise fluctuations .
non - detections in our experiment agree with the lack of consistency between the event rates presented in @xcite and grb event rates .
a joint analysis of our results and four other grb experiments shows that the b12 results are highly unlikely to be astrophysical .
if b12 events are real then the combined analysis constrains the radio spectral index of the events to be @xmath233 .
the international centre for radio astronomy research is a joint venture between curtin university and the university of western australia , funded by the state government of western australia and the joint venture partners .
s.j.t is a western australian premiers research fellow .
r.b.w is supported via the western australian centre of excellence in radio astronomy science and engineering .
the centre for all - sky astrophysics is an australian research council centre of excellence , funded by grant ce110001020 .
d.p is supported by a csirs scholarship provided by curtin university .
j. m. is a super science fellow , supported by an arc grant .
we would like to thank the referee for his or her useful comments and suggestions for this manuscript .
we also thank peter a. curran for useful discussions and suggestion on the x - ray light curves and analysis .
we would also like to thank aidan hotan and bruce stansby for providing the vela data .
this research has made use of the following facilities : * the hobart 26 m radio telescope located at the mount pleasant radio observatory , operated by the university of tasmania ; * the work was supported by ivec through the use of advanced computing resources located at ivec@curtin ; * the swift science data centre located at the university of leicester , uk .
data incoming into the dedisperser are discrete samples in frequency and time .
the de - dispersion process sums the power samples across spectral channels , with the delay for each channel computed according to the frequency and dm ( figure [ dm0 ] ) . under normal circumstances the power level in each spectral channel follows a gaussian distribution . for low dms ,
there is a @xmath050@xmath8 overlap between the power samples that contribute to consecutive dm trials , at the same time .
here we demonstrate that this overlap does not correspond to a high level of correlation in detections between neighbouring dm steps for a large ( e.g @xmath4 ) detection threshold , and therefore that each dm step can be considered independent .
let us consider a dm trial where the power samples in the channels are normally distributed across the spectrum .
the variance of summing @xmath234 independent samples equals @xmath235 and the standard deviation , @xmath236 , where @xmath237 is the noise level in a given power sample .
the probability of obtaining an event above a threshold , @xmath238 , due to random addition of power sample fluctuations is given by the cdf ; @xmath239 , \label{gaussian_cdf}\ ] ] substituting @xmath240 in equation [ gaussian_cdf ] , gives the cdf in terms of the noise level in each power sample ; @xmath241 .
\label{gaussian_cdf_sigma}\ ] ] the probability that we obtain a detection above our experimental threshold ( 5@xmath29 ) from an independent dm step is given by , @xmath242 = \frac{1}{2 } \ , - \ , \frac{1}{2 } \
, \rm erf\left[\frac{5 \ : \sqrt{128 } \ : \sigma_{i}}{\sqrt{2 } \ : \sqrt{128 } \ : \sigma_{i}}\right ] = 2.87 \times 10^{-7}. \label{gaussian_cdf_value}\ ] ] where @xmath243 . for low dms
, we note that due to our choice of progression in the dm steps , there is a @xmath050@xmath8 overlap between the power samples that contribute to consecutive dm trials , at the same time .
when there are shared samples between consecutive dm steps , those samples contribute the same power to both dm trials .
there are many permutations for how power can be distributed in each set of power samples to combine to yield two detections above @xmath244 .
it can be shown that the power distribution with the highest joint probability over the two dm trials corresponds to the case shown in figure [ two_dm ] , where the shared power samples contribute proportionally to the power required to meet the threshold . under this scenario , _ each _ of the three sets of @xmath245 samples contributes [email protected]@xmath247 .
the probability that each set obtains a detection @xmath248 is given by , @xmath249,\ ] ] with @xmath250 , giving , @xmath251.\ ] ] of power samples at dm of 1 overlaps with dm of 0 represented in dotted pink lines . and rest are the non - overlapping samples in the dm of 0 and 1 repersented in dotted black and yellow lines.[two_dm],width=336,height=230 ] when 50@xmath8 of samples overlap and the threshold , @xmath252 , the joint probability of obtaining detections in two consecutive dm trials is given by the product of the probability for each of the three sets ; @xmath253^{3 } = 1.7 \times 10^{-9}.\ ] ] this is @xmath0100 times smaller than the probability of detecting a single event ( equations [ gaussian_cdf]-[gaussian_cdf_value ] ) .
therefore , although the power samples contributing to the consecutive dm trials are 50@xmath8 correlated , due to the improbability of obtaining a single 5@xmath29 detection , the joint probability of obtaining two detections is very small .
we therefore conclude that the dm trials can be considered statistically independent . | the detection of six fast radio bursts ( frbs ) has recently been reported .
frbs are short duration ( @xmath0 1 ms ) , highly dispersed radio pulses from astronomical sources .
the physical interpretation for the frbs remains unclear but is thought to involve highly compact objects at cosmological distance .
it has been suggested that a fraction of frbs could be physically associated with gamma - ray bursts ( grbs ) .
recent radio observations of grbs have reported the detection of two highly dispersed short duration radio pulses using a 12 m radio telescope at 1.4 ghz . motivated by this result ,
we have performed a systematic and sensitive search for frbs associated with grbs . we have observed five grbs at 2.3 ghz using a 26 m radio telescope located at the mount pleasant radio observatory , hobart .
the radio telescope was automated to rapidly respond to gamma - ray coordination network notifications from the swift satellite and slew to the grb position within @xmath0 140 s. the data were searched for pulses up to 5000 pc @xmath1 in dispersion measure and pulse widths ranging from 640 @xmath2s to 25.60 ms .
we did not detect any events @xmath3 .
an in - depth statistical analysis of our data shows that events detected above @xmath4 are consistent with thermal noise fluctuations only .
a joint analysis of our data with previous experiments shows that previously claimed detections of frbs from grbs are unlikely to be astrophysical .
our results are in line with the lack of consistency noted between the recently presented frb event rates and grb event rates . |
providing an explicit definition of time is an extremely difficult endeavor , although it does seem to be intimately related to change , an idea reflected in aristotle s famous metaphor : _ time is the moving image of eternity _ @xcite . in fact , one may encounter many reflections and philosophical considerations on time over the ages , culminating in newton s notion of absolute time .
newton stated that time flowed at the same rate for all observers in the universe .
but , in 1905 , einstein changed altogether our notion of time .
time flowed at different rates for different observers , and minkowski , three years later , formally united the parameters of time and space , giving rise to the notion of a four - dimensional entity , spacetime . adopting a pragmatic point of view , this assumption seems reasonable , as to measure time a changing configuration of matter is needed , i.e. , a swinging pendulum , etc .
change seems to be imperative to have an emergent notion of time .
therefore , time is empirically related to change . but change can be considered as a variation or sequence of occurrences @xcite .
thus , intuitively , a sequence of successive occurrences provides us with a notion of something that flows , i.e. , it provides us with the notion of _ time_. time flows and everything relentlessly moves along this stream . in relativity , we can substitute the above empirical notion of a sequence of occurrences by a sequence of _
we idealize the concept of an event to become a point in space and an instant in time . following this reasoning of thought , a sequence of events has a determined _ temporal order_. we experimentally verify that specific events occur before others and not vice - versa .
certain events ( effects ) are triggered off by others ( causes ) , providing us with the notion of _
causality_. thus , the conceptual definition and understanding of time , both quantitatively and qualitatively is of the utmost difficulty and importance .
special relativity provides us with important quantitative elucidations of the fundamental processes related to time dilation effects .
the general theory of relativity ( gtr ) provides a deep analysis to effects of time flow in the presence of strong and weak gravitational fields @xcite . as time
is incorporated into the proper structure of the fabric of spacetime , it is interesting to note that gtr is contaminated with non - trivial geometries which generate _ closed timelike curves _
a closed timelike curve ( ctc ) allows time travel , in the sense that an observer which travels on a trajectory in spacetime along this curve , returns to an event which coincides with the departure .
the arrow of time leads forward , as measured locally by the observer , but globally he / she may return to an event in the past .
this fact apparently violates causality , opening pandora s box and producing time travel paradoxes @xcite , throwing a veil over our understanding of the fundamental nature of time .
the notion of causality is fundamental in the construction of physical theories , therefore time travel and it s associated paradoxes have to be treated with great caution .
the paradoxes fall into two broad groups , namely the _ consistency paradoxes _ and the _ causal loops_. the consistency paradoxes include the classical grandfather paradox .
imagine traveling into the past and meeting one s grandfather . nurturing homicidal tendencies , the time traveler murders his grandfather , impeding the birth of his father , therefore making his own birth impossible .
in fact , there are many versions of the grandfather paradox , limited only by one s imagination .
the consistency paradoxes occur whenever possibilities of changing events in the past arise .
the paradoxes associated to causal loops are related to self - existing information or objects , trapped in spacetime .
imagine a time traveler going back to his past , handing his younger self a manual for the construction of a time machine .
the younger version then constructs the time machine over the years , and eventually goes back to the past to give the manual to his younger self .
the time machine exists in the future because it was constructed in the past by the younger version of the time traveler .
the construction of the time machine was possible because the manual was received from the future .
both parts considered by themselves are consistent , and the paradox appears when considered as a whole .
one is liable to ask , what is the origin of the manual , for it apparently surges out of nowhere .
there is a manual never created , nevertheless existing in spacetime , although there are no causality violations .
an interesting variety of these causal loops was explored by gott and li @xcite , where they analyzed the idea of whether there is anything in the laws of physics that would prevent the universe from creating itself .
thus , tracing backwards in time through the original inflationary state a region of ctcs may be encountered , giving _ no _ first - cause .
a great variety of solutions to the einstein field equations ( efes ) containing ctcs exist , but , two particularly notorious features seem to stand out .
solutions with a tipping over of the light cones due to a rotation about a cylindrically symmetric axis ; and solutions that violate the energy conditions of gtr , which are fundamental in the singularity theorems and theorems of classical black hole thermodynamics @xcite .
a great deal of attention has also been paid to the quantum aspects of closed timelike curves @xcite . throughout this paper , we use the notation @xmath0 .
it is interesting to note that the tipping over of light cones seems to be a generic feature of some solutions with a rotating cylindrical symmetry .
the general metric for a stationary , axisymmetric solution with rotation is given by @xcite @xmath1 where @xmath2 is the distance along the axis of rotation ; @xmath3 is the angular coordinate ; @xmath4 is the radial coordinate ; and @xmath5 is the temporal coordinate .
the metric components are only functions of the radial coordinate @xmath4 .
note that the determinant , @xmath6 , is lorentzian provided that @xmath7 . due to the periodic nature of the angular coordinate , @xmath3 , an azimuthal curve with @xmath8 is a closed curve of invariant length @xmath9 .
if @xmath10 is negative then the integral curve with @xmath11 fixed is a ctc . if @xmath12 , then the azimuthal curve is a closed null curve .
now , consider a null azimuthal curve , not necessarily a geodesic nor closed , in the @xmath13 plane with @xmath14 fixed .
the null condition , @xmath15 , implies @xmath16 with @xmath17 . solving the quadratic , we have @xmath18 due to the lorentzian signature constraint , @xmath19 , the roots are real .
if @xmath20 then the light cones are tipped over sufficiently far to permit a trip to the past . by going once around the azimuthal direction ,
the total backward time - jump for a null curve is @xmath21 if @xmath20 for even a single value of @xmath4 , the chronology - violation region covers the entire spacetime @xcite .
thus , the tilting of light cones are generic features of spacetimes which contain ctcs , as depicted in fig . [ tipcones ] .
the present section is far from making an exhaustive search of all the efe solutions generating ctcs with these features , but the best known spacetimes will be briefly analyzed , namely , the van stockum spacetime , the gdel universe , the spinning cosmic strings and the gott two - string time machine , which is a variation on the theme of the spinning cosmic string . the earliest solution to the efes containing ctcs , is probably that of the van stockum spacetime , which describes a stationary , cylindrically symmetric solution of a rapidly rotating infinite cylinder of dust , surrounded by vacuum .
the centrifugal forces of the dust are balanced by the gravitational attraction .
the metric , assuming the respective symmetries , takes the form of eq .
( [ stationarymetric ] ) , and @xmath5 is required to be timelike at @xmath22 .
the coordinates @xmath23 have the following domain @xmath24 the metric for the interior solution @xmath25 , where @xmath26 is the surface of the cylinder , is given by @xmath27 where @xmath28 is the angular velocity of the cylinder .
it is immediate to verify that ctcs arise if @xmath29 , i.e. , for @xmath30 the azimuthal curves with @xmath11 fixed are ctcs .
the condition @xmath31 is imposed .
the causality violation region could be eliminated by requiring that boundary of the cylinder to be at @xmath32 .
the interior solution would then be joined to an exterior solution , which would be causally well - behaved .
the resulting upper bound to the `` velocity '' @xmath33 would be @xmath34 , although the orbits of the particles creating the field are timelike for all @xmath4 . applying the efe ,
the energy density and @xmath35-velocity of the dust are given by @xmath36 respectively .
the coordinate system co - rotates with the dust .
the source is simply positive density dust , implying that all of the energy condition are satisfied .
van stockum developed a procedure which generates an exterior solution for all @xmath37 @xcite .
consider the following range : 1 .
@xmath38 @xmath39 .
+ the exterior solution is given by the following functions @xmath40 with @xmath41 2 .
@xmath42 @xmath43 .
@xmath44,\\ & & m(r)=(r/2 ) \left [ 1+\ln \left(r / r \right ) \right ] , % \qquad % f(r)=(r / r ) \left [ 1-\ln \left(r / r \right ) \right ] \,.\end{aligned}\ ] ] 3 .
@xmath45 @xmath46 .
@xmath47 with @xmath48 as in the interior solution , @xmath49 , so that the metric signature is lorentzian for @xmath50 .
one may show that the causality violation is avoided for @xmath51 , but in the region @xmath52 , ctcs appear .
the causality violations arise from the sinusoidal factors of the metric components .
the first zero of @xmath10 occurs at @xmath53 \,.\ ] ] thus , causality violation occur in the matter - free space surrounding a rapidly rotating infinite cylinder , as shown in figure [ fig : stockum ] .
the van stockum spacetime is not asymptotically flat .
but , the gravitational potential of the cylinder s newtonian analog also diverges at radial infinity . shrinking the cylinder down to a `` ring '' singularity ,
one ends up with the kerr solution , which also has ctcs ( the causal structure of the kerr spacetime has been extensively analyzed by de felice and collaborators @xcite ) . in summary ,
the van stockum solution contains ctc provided @xmath54 .
the chronology - violating region covers the entire spacetime .
reactions to the van stockum solution is that it is unphysical , as it applies to an infinitely long cylinder and it is not asymptotically flat .
kurt gdel in @xmath55 discovered an exact solution to the efes of a uniformly rotating universe containing dust and a nonzero cosmological constant @xcite .
the total energy - momentum is given by @xmath56 however , the latter may be expressed in terms of a perfect fluid , with rotation , energy density @xmath57 and pressure @xmath58 , in a universe with a zero cosmological constant , i.e. , @xmath59 with the following definitions @xmath60 the manifold is @xmath61 and the metric of the gdel solution is provided by @xmath62 the four - velocity and the vorticity of the fluid are , @xmath63 and @xmath64 , respectively .
the einstein field equations provide the following stress - energy scenario : @xmath65 thus , the null , weak and dominant energy conditions are satisfied , while the dominant energy condition is in the imminence of being violated .
note that the metric ( [ godelmetric ] ) is the direct sum of the metrics @xmath66 and @xmath67 .
the metric @xmath66 is given by @xmath68 with the manifold @xmath69 defined by the coordinates @xmath70 .
the metric @xmath67 is given by @xmath71 , with the manifold @xmath72 , defined by the coordinate @xmath2 . to analyze the causal properties of the solution
, it is sufficient to consider @xmath73 .
consider a set of alternative coordinates @xmath74 in @xmath73 , in which the rotational symmetry of the solution , around the axis @xmath22 , is manifest and suppressing the irrelevant @xmath2 coordinate , defined by @xcite @xmath75&= & e^{-2r}\,\tan(\phi/2 ) \nonumber \,,\end{aligned}\ ] ] so that the metric ( [ godelmetric1 ] ) takes the form @xmath76 \,.\ ] ] moving away from the axis , the light cones open out and tilt in the @xmath3-direction .
the azimuthal curves with @xmath77 are ctcs if the condition @xmath78 is satisfied .
it is interesting to note that in the gdel spacetime , closed timelike curves are not geodesics .
however , novello and rebouas @xcite discovered a new generalized solution of the gdel metric , of a shear - free nonexpanding rotating fluid , in which successive concentric causal and noncausal regions exist , with closed timelike curves which are geodesics .
a complete study of geodesic motion in gdel s universe , using the method of the effective potential was further explored by novello _
et al _ @xcite .
much interest has been aroused in time travel in the gdel spacetime , from which we may mention the analysis of the geodesical and non - geodesical motions considered by pfarr @xcite and malament @xcite .
the string spacetime is assumed to be static and cylindrically symmetric , with the string lying along the axis of symmetry .
the most general static , cylindrically symmetric metric has the form @xmath79 where @xmath80 , @xmath81 and @xmath82 are functions of @xmath83 .
@xmath84 and @xmath85 are identified .
suppose that the string has a uniform density @xmath86 , out to some cylindrical radius @xmath87 .
the end results will prove to be independent of @xmath87 , so that the string s transverse dimensions may be reduced to zero , yielding an unambiguous exact exterior metric for the string .
the stress - energy tensor of the string , in an orthonormal frame , is given by @xmath88 and all the other components are equal to zero , for @xmath89 @xcite . the resulting efes are given by : @xmath90 = 8\pi \,\epsilon \ , , \label{stringefett } \\ & & e^{-2\lambda}\,\left(\lambda'\,\phi'+\nu'\,\lambda'+\lambda'\,\phi ' \right ) = 0 \ , , \label{stringeferr } \\ & & e^{-2\lambda}\,\left[\lambda''+\nu''+(\nu')^2 \right ] = 0 \ , , \label{stringefethetatheta } \\ & & e^{-2\lambda}\,\left[-\lambda'\,\phi'-\nu'\,\lambda'+\phi '' + ( \phi')^2+\nu''+(\nu')^2+\nu'\,\phi ' \right ] = -8\pi\,\epsilon \ , , \label{stringefezz}\end{aligned}\ ] ] where the prime denotes a derivative with respect to @xmath83 .
these are non - linear equations for the metric functions , and are easily solved in the case of the uniform density string .
conservation of the stress - energy , @xmath91 , yields @xmath92 this implies that through eq .
( [ stringefethetatheta ] ) , @xmath80 and @xmath82 are constant , and may be set to zero by an appropriate rescaling of the coordinates @xmath5 , @xmath83 and @xmath2 .
equation ( [ stringeferr ] ) is then satisfied automatically and eqs .
( [ stringefett ] ) and ( [ stringefezz ] ) become identical , i.e. , @xmath93 substituting @xmath94 , i.e. , @xmath95 , yields @xmath96 where @xmath97 .
the metric on the axis will be flat , i.e. , no cone singularity , if @xmath98 and @xmath99 .
thus , the interior metric of a uniform - density string is then given by @xmath100 the exterior metric for the string spacetime must be a static , cylindrically symmetric , vacuum solution of the efes .
the most general solution , discovered by levi - civita @xcite is given @xmath101 \label{extstring}\,,\ ] ] where @xmath102 and @xmath103 are freely chosen constants .
the string is lorentz invariant in the @xmath2-direction .
requiring that the metric ( [ extstring ] ) be lorentz invariant in the @xmath2-direction restricts the values of @xmath102 , namely , @xmath104 and @xmath105 @xcite .
one may now join the interior and exterior metrics together along the surface of the string at @xmath106 and @xmath107 .
the darmois - israel junction conditions require that the intrinsic metrics induced on the junction surface by the interior and exterior metrics be identical , and that the discontinuity in the extrinsic curvature of the surface be related to the surface stress - energy .
consider the @xmath104 flat exterior case .
the intrinsic metric can then be matched by requiring @xmath108 , @xmath109 and @xmath110 .
the latter condition provides @xmath111 calculating the extrinsic curvature tensors and equating them to each other , so as to have no surface stress - energy present , one obtains the following relation @xmath112 combining this with the intrinsic metric constraint , eq .
( [ stringintrinsic ] ) , to eliminate @xmath113 , yields @xmath114 the exterior metric of the string is then given by eq .
( [ extstring ] ) with @xmath104 , and @xmath103 given by eq .
( [ def : a ] ) .
the concept of a mass per unit length for a cylindrically symmetric source in general relativity is not unambiguously defined , unlike the case of spherical symmetry . for a static ,
cylindrically symmetric spacetime , a useful simple definition is to integrate the energy - density , @xmath115 over the proper volume of the source , i.e. , the string .
the mass per unit length , or linear energy - density , is given by @xmath116 \,,\ ] ] which , taking into account @xmath97 , reduces to @xmath117 thus , the exact exterior metric is given by @xmath118 which will be used below in the gott cosmic string spacetime .
an extremely elegant model of a time - machine was constructed by gott @xcite .
the gott time - machine is an exact solution of the efe for the general case of two moving straight cosmic strings that do not intersect @xcite .
this solution produces ctcs even though they do not violate the wec , have no singularities and event horizons , and are not topologically multiply - connected as the wormhole solution ( see below ) .
the appearance of ctcs relies solely on the gravitational lens effect and the relativity of simultaneity .
we follow the analysis of ref .
@xcite closely throughout this section .
the exterior metric of a straight cosmic string is given by eq .
( [ exactextstring ] ) .
the geometry of a @xmath119 , @xmath120 section of this solution is that of a cone with an angle deficit @xmath121 in the exterior ( vacuum ) region . applying a new coordinate @xmath122 ,
the exterior metric becomes @xmath123 where @xmath124 .
the above metric is the metric for minkowski space in cylindrical coordinates where a wedge of angle deficit @xmath121 is missing , and points with coordinates @xmath125 and @xmath126 are identified .
now , the static solution for two parallel cosmic strings separated by a distance @xmath127 is constructed in the following manner .
consider the metric ( [ exactextstring ] ) , by replacing the angular and radial coordinates , @xmath128 and @xmath4 , respectively by the cartesian coordinates , @xmath129 and @xmath130 .
this reduces the metric to @xmath131 , with the following restrictions : @xmath132 and the points with @xmath133 are identified .
the @xmath134-surface @xmath135 , has the metric @xmath136 with zero intrinsic and extrinsic curvature , as it is part of a @xmath137-dimensional minkowski spacetime .
it is thus possible to produce a mirror - image copy of the region @xmath138 , including the interior solution , by joining it along the three - surface @xmath135 .
this second solution lies in the @xmath139 region .
the two copies obey all the matching conditions along the surface @xmath135 , because the latter is a @xmath140-dimensional minkowskian spacetime with zero intrinsic and extrinsic curvature .
see figure [ gotttm ] for details .
plane , width=326 ] consider now two observers @xmath141 and @xmath142 at rest with respect to the cosmic strings with world lines given by @xmath143 respectively .
it is possible to prove that observer @xmath142 sees three images of the observer @xmath141 @xcite .
the central image is from a geodesic passing through the origin @xmath144 , the two outer images , which are displaced from the central image by an angle @xmath145 on each side , represent geodesics that pass through events @xmath146 and @xmath147 .
note that @xmath148 and @xmath149 are identified , as are @xmath150 and @xmath151 .
considering the following trigonometric relationship @xmath152 it is simple to verify that the value of @xmath153 to minimize @xmath154 is @xmath155 .
thus , we have @xmath156 if @xmath157 , and the light beam going through @xmath144 with a gravitational lensing time delay between the two images of @xmath158 .
note that if a light beam traversing through @xmath146 can beat a light beam traveling through @xmath144 , then so can a spaceship traveling at a high enough velocity , @xmath159 , relative to the string .
the spaceship connects two events in the @xmath135 @xmath140-dimensional minkowski spacetime with a spacelike separation .
let the spaceship begin at @xmath141 and end at @xmath142 , given by the following events @xmath160 respectively .
the time for the spaceship to traverse from @xmath161 , through @xmath146 to @xmath162 is @xmath163 .
the separation of @xmath161 and @xmath164 is spacelike providing that @xmath165 , which can always be verified for high enough @xmath166 , for @xmath156 .
the following step is to give the @xmath138 solution a boost with velocity @xmath167 in the @xmath168-direction via a lorentz transformation such that @xmath161 and @xmath162 become simultaneous in the laboratory frame .
the velocity for the simultaneity to occur is @xmath169 .
analogously , we give the @xmath170 solution a boost with velocity @xmath167 in the @xmath171-direction .
the two solutions @xmath172 and @xmath170 may still be matched together because the lorentz transformations do not alter the fact that the boundary surface @xmath173 in each solution is still a @xmath140-dimensional minkowskian spacetime with zero intrinsic and extrinsic curvature .
the spaceship goes from @xmath161 through @xmath146 and arrives at @xmath162 , which is simultaneous in the laboratory frame . by symmetry
, the spaceship travels in the opposite direction past the oppositely moving string through @xmath147 and arrives back at event @xmath161 , which is also simultaneous with @xmath162 in the laboratory frame .
the spaceship has completed a ctc , as it encircles the two parallel cosmic strings as they pass each other in a sense opposite to that of the strings relative motion . in principle , it is also possible to find a reference frame in which the spaceship arrives at @xmath161 before it s departure .
the events in the laboratory frame have the following coordinates : @xmath174 and @xmath175 with @xmath176 and since @xmath166 , we have @xmath177 or @xmath178 considering the following approximations , @xmath179 , we have @xmath180 or simply @xmath181 for @xmath182 expected for grand unified cosmic strings , we have @xmath183 in order to produce ctcs . in the laboratory frame
it is clear how the ctc is created .
the @xmath146 and @xmath147 identifications allow the particle to effectively travel backward in time twice in the laboratory frame .
the identifications of @xmath146 and @xmath147 is equivalent to having a complete minkowski spacetime without the missing wedges where instantaneous , tachyonic , travel in the string rest frames between @xmath148 and @xmath149 , @xmath150 and @xmath151 , is possible .
it is also interesting to verify whether the ctcs in the gott solution appear at some particular moment , i.e. , when the strings approach each other s neighborhood , or if they already pre - exist , i.e. , they intersect any spacelike hypersurface .
these questions are particularly important in view of hawking s chronology protection conjecture @xcite .
this conjecture states that the laws of physics prevent the creation of ctcs .
if correct , then the solutions of the efe which admit ctcs are either unrealistic or are solutions in which the ctcs are pre - existing , so that the time -machine is not created by dynamical processes .
amos ori proved that in gott s spacetime , ctcs intersect every @xmath184 hypersurface @xcite , so that it is not a counter - example to the chronology protection conjecture .
the global structure of the gott spacetime was further explored by cutler @xcite , and it was shown that the closed timelike curves are confined to a certain region of the spacetime , and that the spacetime contains complete spacelike and achronal hypersurfaces from which the causality violating regions evolve .
grant also examined the global structure of the two - string spacetime and found that away from the strings , the space is identical to a generalized misner space @xcite .
the vacuum expectation value of the energy - momentum tensor for a conformally coupled scalar field was then calculated on the respective generalized misner space , which was found to diverge weakly on the chronology horizon , but diverge strongly on the polarized hypersurfaces .
thus , the back reaction due to the divergent behavior around the polarized hypersurfaces are expected to radically alter the structure of spacetime , before quantum gravitational effects become important , suggesting that hawking s chronology protection conjecture holds for spaces with a noncompactly generated chronology horizon .
soon after , laurence @xcite showed that the region containing ctcs in gott s two - string spacetime is identical to the regions of the generalized misner space found by grant , and constructed a family of isometries between both gott s and grant s regions .
this result was used to argue that the slowly diverging vacuum polarization at the chronology horizon of the grant space carries over without change to the gott space .
furthermore , it was shown that the gott time machine is unphysical in nature , for such an acausal behavior can not be realized by physical and timelike sources @xcite .
consider an infinitely long straight string that lies and spins around the @xmath2-axis .
the symmetries are analogous to the van stockum spacetime , but the asymptotic behavior is different @xcite .
we restrict the analysis to an infinitely long straight string , with a delta - function source confined to the @xmath2-axis .
it is characterized by a mass per unit length , @xmath185 ; a tension , @xmath186 , and an angular momentum per unit length , @xmath187 . for cosmic strings ,
the mass per unit length is equal to the tension , @xmath188 . in cylindrical coordinates the metric takes the following form @xmath189 ^ 2+dr^2+(1 - 4\mu)^2\,r^2\;d\varphi^2+dz^2 \,,\ ] ] with the following coordinate range @xmath190 adopting a new set of coordinates @xmath191 the metric may be rewritten as @xmath192 with a new coordinate range @xmath193 subject to the following identifications @xmath194 \,.\ ] ] outside the core @xmath22 , the metric is locally flat , i.e. , the riemann tensor is zero .
the geometry is that of flat minkowski spacetime subject to a somewhat peculiar set of identifications . on traveling once around the the string
, one sees that the spatial slices are `` missing '' a wedge of angle @xmath195 , which defines the deficit angle @xmath196 . on traveling once around the string ,
one undergoes a backward time - jump of @xmath197 consider an azimuthal curve , i.e. , an integral curve of @xmath198 .
closed timelike curves appear whenever @xmath199 these ctcs can be deformed to cover the entire spacetime , consequently , the chronology - violating region covers the entire manifold .
the traditional manner of solving the efes , @xmath200 , consists in considering a plausible stress - energy tensor , @xmath201 , and finding the geometrical structure , @xmath202 .
but one can run the efe in the reverse direction by imposing an exotic metric @xmath203 , and eventually finding the matter source for the respective geometry . in this fashion ,
solutions violating the energy conditions have been obtained . adopting the reverse philosophy , solutions such as traversable wormholes , the warp drive , the krasnikov tube and the ori - soen spacetime
have been obtained .
these solutions violate the energy conditions and with simple manipulations generate ctcs .
much interest has been aroused in traversable wormholes since the classical article by morris and thorne @xcite .
a wormhole is a hypothetical tunnel which connects different regions in spacetime .
these solutions are multiply - connected and probably involve a topology change , which by itself is a problematic issue .
consider the following spherically symmetric and static wormhole solution @xmath204 where @xmath205 and @xmath206 are arbitrary functions of the radial coordinate @xmath4 .
@xmath205 is denoted the redshift function , for it is related to the gravitational redshift , and @xmath206 is denoted the shape function , because as can be shown by embedding diagrams , it determines the shape of the wormhole @xcite .
the coordinate @xmath4 is non - monotonic in that it decreases from @xmath207 to a minimum value @xmath113 , representing the location of the throat of the wormhole , where @xmath208 , and then it increases from @xmath113 to @xmath207 .
a fundamental property of a wormhole is that a flaring out condition of the throat , given by @xmath209 , is imposed @xcite , and at the throat @xmath210 , the condition @xmath211 is imposed to have wormhole solutions .
it is precisely these restrictions that impose the nec violation in classical general relativity .
another condition that needs to be satisfied is @xmath212 . for the wormhole to be traversable
, one must demand that there are no horizons present , which are identified as the surfaces with @xmath213 , so that @xmath205 must be finite everywhere .
several candidates have been proposed in the literature , amongst which we refer to solutions in higher dimensions , for instance in einstein - gauss - bonnet theory @xcite , wormholes on the brane @xcite ; solutions in brans - dicke theory @xcite ; wormholes constructed in @xmath214 gravity @xcite ; wormhole solutions in semi - classical gravity ( see ref . @xcite and references therein ) ; exact wormhole solutions using a more systematic geometric approach were found @xcite ; wormhole solutions and thin shells @xcite ; geometries supported by equations of state responsible for the cosmic acceleration @xcite ; spherical wormholes were also formulated as an initial value problem with the throat serving as an initial value surface @xcite ; solutions in conformal weyl gravity were found @xcite , and thin accretion disk observational signatures were also explored @xcite , etc ( see refs .
@xcite for more details and @xcite for a recent review ) .
one of the most fascinating aspects of wormholes is their apparent ease in generating ctcs @xcite .
there are several ways to generate a time machine using multiple wormholes @xcite , but a manipulation of a single wormhole seems to be the simplest way @xcite .
the basic idea is to create a time shift between both mouths .
this is done invoking the time dilation effects in special relativity or in general relativity , i.e. , one may consider the analogue of the twin paradox , in which the mouths are moving one with respect to the other , or simply the case in which one of the mouths is placed in a strong gravitational field . to create a time shift using the twin paradox analogue ,
consider that the mouths of the wormhole may be moving one with respect to the other in external space , without significant changes of the internal geometry of the handle .
for simplicity , consider that one of the mouths @xmath141 is at rest in an inertial frame , whilst the other mouth @xmath142 , initially at rest practically close by to @xmath141 , starts to move out with a high velocity , then returns to its starting point .
due to the lorentz time contraction , the time interval between these two events , @xmath215 , measured by a clock comoving with @xmath142 can be made to be significantly shorter than the time interval between the same two events , @xmath216 , as measured by a clock resting at @xmath141 .
thus , the clock that has moved has been slowed by @xmath217 relative to the standard inertial clock .
note that the tunnel ( handle ) , between @xmath141 and @xmath142 remains practically unchanged , so that an observer comparing the time of the clocks through the handle will measure an identical time , as the mouths are at rest with respect to one another .
however , by comparing the time of the clocks in external space , he will verify that their time shift is precisely @xmath218 , as both mouths are in different reference frames , frames that moved with high velocities with respect to one another .
now , consider an observer starting off from @xmath141 at an instant @xmath219 , measured by the clock stationed at @xmath141 .
he makes his way to @xmath142 in external space and enters the tunnel from @xmath142 .
consider , for simplicity , that the trip through the wormhole tunnel is instantaneous .
he then exits from the wormhole mouth @xmath141 into external space at the instant @xmath220 as measured by a clock positioned at @xmath141 .
his arrival at @xmath141 precedes his departure , and the wormhole has been converted into a time machine .
see figure [ fig : wh - time - machine ]
. for concreteness , following the morris _
et al _ analysis @xcite , consider the metric of the accelerating wormhole given by @xmath221 where the proper radial distance , @xmath222 , is used .
@xmath223 is a form function that vanishes at the wormhole mouth @xmath141 , at @xmath224 , rising smoothly from 0 to 1 , as one moves to mouth @xmath142 ; @xmath225 is the acceleration of mouth @xmath142 as measured in its own asymptotic rest frame .
consider that the external metric to the respective wormhole mouths is @xmath226 .
thus , the transformation from the wormhole mouth coordinates to the external lorentz coordinates is given by @xmath227 for mouth @xmath141 , where @xmath228 is the time - independent @xmath229 location of the wormhole mouth @xmath141 , and @xmath230 for the accelerating wormhole mouth @xmath142 .
the world line of the center of mouth @xmath142 is given by @xmath231 and @xmath232 with @xmath233 ; @xmath234 is the velocity of mouth @xmath142 and @xmath235 the respective lorentz factor ; the acceleration appearing in the wormhole metric is given @xmath236 @xcite .
novikov considered other variants of inducing a time shift through the time dilation effects in special relativity , by using a modified form of the metric ( [ accerelatedwh ] ) , and by considering a circular motion of one of the mouths with respect to the other @xcite .
another interesting manner to induce a time shift between both mouths is simply to place one of the mouths in a strong external gravitational field , so that times slows down in the respective mouth .
the time shift will be given by @xmath237 @xcite .
at the wormhole throat is marked off , and note that identical values are the same event as seen through the wormhole handle . in figure @xmath238 , mouth
@xmath141 remains at rest , while mouth @xmath142 accelerates from @xmath141 at a high velocity , then returns to its starting point at rest .
a time shift is induced between both mouths , due to the time dilation effects of special relativity .
the light cone - like hypersurface @xmath239 shown is a cauchy horizon . through every event to the future of @xmath239 there
exist ctcs , and on the other hand there are no ctcs to the past of @xmath239 . in figure @xmath240 ,
a time shift between both mouths is induced by placing mouth @xmath142 in strong gravitational field .
see text for details.,title="fig:",width=249 ] at the wormhole throat is marked off , and note that identical values are the same event as seen through the wormhole handle . in figure @xmath238 , mouth
@xmath141 remains at rest , while mouth @xmath142 accelerates from @xmath141 at a high velocity , then returns to its starting point at rest .
a time shift is induced between both mouths , due to the time dilation effects of special relativity .
the light cone - like hypersurface @xmath239 shown is a cauchy horizon . through every event to the future of @xmath239 there
exist ctcs , and on the other hand there are no ctcs to the past of @xmath239 . in figure @xmath240 ,
a time shift between both mouths is induced by placing mouth @xmath142 in strong gravitational field .
see text for details.,title="fig:",width=230 ] a time - machine model was also proposed by amos ori and yoav soen which significantly ameliorates the conditions of the efe s solutions which generate ctcs @xcite .
the ori - soen model presents some notable features .
it was verified that ctcs evolve , within a bounded region of space , from a well - defined initial slice @xmath241 , a partial cauchy surface , which does not display causality violation .
the partial cauchy surface and spacetime are asymptotically flat , contrary to the gott spacetime , and topologically trivial , contrary to the wormhole solutions .
the causality violation region is constrained within a bounded region of space , and not in infinity as in the gott solution .
the wec is satisfied up until and beyond a time slice @xmath242 , on which the ctcs appear .
more recently , ori presented a class of curved - spacetime vacuum solutions which develop closed timelike curves at some particular moment @xcite .
these vacuum solutions were then used to construct a time - machine model .
the causality violation occurs inside an empty torus , which constitutes the time - machine core .
the matter field surrounding this empty torus satisfies the weak , dominant , and strong energy conditions .
the model is regular , asymptotically flat , and topologically trivial , although stability still remains the main open question . within the framework of general relativity ,
it is possible to warp spacetime in a small _ bubblelike _ region @xcite , in such a way that the bubble may attain arbitrarily large velocities , @xmath243 . inspired in the inflationary phase of the early universe , the enormous speed of separation arises from the expansion of spacetime itself .
the model for hyperfast travel is to create a local distortion of spacetime , producing an expansion behind the bubble , and an opposite contraction ahead of it ( see also @xcite ) . in the alcubierre warp
drive the spacetime metric is @xmath244 ^ 2 \label{cartesianwarpmetric}\,.\ ] ] the form function @xmath245 possesses the general features of having the value @xmath246 in the exterior and @xmath247 in the interior of the bubble .
the general class of form functions , @xmath245 , chosen by alcubierre was spherically symmetric : @xmath248 with @xmath249 .
then @xmath250 ^ 2+x^2+y^2\right\}^{1/2}.\ ] ] consider the following form @xmath251 -\tanh\left[\sigma(r - r)\right]}{2\tanh(\sigma r)}\ , , \label{e : form}\ ] ] in which @xmath252 and @xmath253 are two arbitrary parameters .
@xmath26 is the `` radius '' of the warp - bubble , and @xmath254 can be interpreted as being inversely proportional to the bubble wall thickness .
if @xmath254 is large , the form function rapidly approaches a _ top hat _ function , i.e. , @xmath255,\\ 0 , & { \rm if}\ ; r\in(r,\infty ) .
\end{array } \right.\ ] ] it can be shown that observers with the four velocity @xmath256 move along geodesics , as their @xmath35-acceleration is zero , _
i.e. _ , @xmath257 . the spaceship , which in the original formulation is treated as a test particle which moves along the curve @xmath258 , can easily be seen to always move along a timelike curve , regardless of the value of @xmath243 .
one can also verify that the proper time along this curve equals the coordinate time , by simply substituting @xmath258 in eq .
( [ cartesianwarpmetric ] ) .
this reduces to @xmath259 , taking into account @xmath260 and @xmath261 .
consider a spaceship placed within the alcubierre warp bubble .
the expansion of the volume elements , @xmath262 , is given by @xmath263 .
taking into account eq .
( [ e : form ] ) , we have ( for alcubierre s version of the warp bubble ) @xmath264 the center of the perturbation corresponds to the spaceship s position @xmath265 .
the volume elements are expanding behind the spaceship , and contracting in front of it , as shown in figure [ alcubierre - expansion ] .
one may consider a hypothetical spaceship immersed within the bubble , moving along a timelike curve , regardless of the value of @xmath243 . due to the arbitrary value of the warp bubble velocity , the metric of the warp drive permits superluminal travel , which raises the possibility of the existence of ctcs .
although the solution deduced by alcubierre by itself does not possess ctcs , everett demonstrated that these are created by a simple modification of the alcubierre metric @xcite , by applying a similar analysis as in tachyons .
the modified metric takes the form @xmath266 with @xmath267^{1/2 } \,.\end{aligned}\ ] ] the spacetime is flat in the exterior of a warp bubble with radius @xmath26 , but now in the modified version is centered in @xmath268 .
the bubble moves with a velocity @xmath269 , on a trajectory parallel with the @xmath2-axis .
one may for simplicity consider the form function given by eq .
( [ e : form ] ) . we shall also impose that @xmath270 , so that the form function is negligible , i.e. , @xmath271 .
now , consider two stars , @xmath272 and @xmath273 , at rest in the coordinate system of the metric ( [ modwarpmetric ] ) , and located on the @xmath2-axis at @xmath173 and @xmath274 , respectively . the metric along the @xmath2-axis is minkowskian as @xmath270 .
therefore , a light beam emitted at @xmath272 , at @xmath173 , moving along the @xmath2-axis with @xmath275 , arrives at @xmath273 at @xmath274 .
suppose that the spaceship initially starts off from @xmath272 , with @xmath276 , moving off to a distance @xmath153 along the @xmath277axis and neglecting the time it needs to cover @xmath135 to @xmath278 .
at @xmath153 , it is then subject to a uniform acceleration , @xmath103 , along the the @xmath279axis for @xmath280 , and @xmath281 for @xmath282 .
the spaceship will arrive at the spacetime event @xmath273 with coordinates @xmath283 and @xmath284 .
once again , the time required to travel from @xmath278 to @xmath135 is negligible . the separation between the two events , departure and arrival is @xmath285 and will be spatial if the following condition is verified @xmath286 in this case , the spaceship will arrive at @xmath273 before the light beam , if the latter s trajectory is a straight line , and both departures are simultaneous from @xmath272 .
inertial observers situated in the exterior of the spaceship , at @xmath272 and @xmath273 , will consider the spaceship s movement as superluminal , since the distance @xmath287 is covered in an interval @xmath288 .
however , the spaceship s wordline is contained within it s light cone .
the worldline of the spaceship is given by @xmath289 , while it s future light cone is given by @xmath290 .
the latter relation can easily be inferred from the null condition , @xmath15 .
since the quadri - vector with components @xmath291 is spatial , the temporal order of the events , departure and arrival , is not well - defined . introducing new coordinates , @xmath292 , obtained by a lorentz transformation , with a boost @xmath293 along the @xmath2-axis .
the arrival at @xmath273 in the @xmath292 coordinates correspond to @xmath294 with @xmath295 .
the events , departure and arrival , will be simultaneous if @xmath296
. the arrival will occur before the departure if @xmath297 , i.e. , @xmath298 the fact that the spaceship arrives at @xmath273 with @xmath299 , does not by itself generate ctcs .
consider the metric , eq .
( [ modwarpmetric ] ) , substituting @xmath2 and @xmath5 by @xmath300 and @xmath301 , respectively ; @xmath243 by @xmath302 ; @xmath103 by @xmath281 ; and @xmath153 by @xmath303 .
this new metric describes a spacetime in which an alcubierre bubble is created at @xmath304 , which moves along @xmath305 and @xmath306 , from @xmath272 to @xmath273 with a velocity @xmath307 , and subject to an acceleration @xmath308 . for observers at rest relatively to the coordinates @xmath292 , situated in the exterior of the second bubble , it is identical to the bubble defined by metric , eq . ( [ modwarpmetric ] ) , as it is seen by inertial observers at rest at @xmath272 and @xmath273 .
the only differences reside in a change of the origin , direction of movement and possibly of the value of acceleration .
the stars , @xmath272 and @xmath273 , are st rest in the coordinate system of the metric , eq .
( [ modwarpmetric ] ) , and in movement along the negative direction of the @xmath2-axis with velocity @xmath293 , relatively to the coordinates @xmath292 .
the two coordinate systems are equivalent due to the lorentz invariance , so if the first is physically realizable , then so is the second . in the new metric , by analogy with eq .
( [ modwarpmetric ] ) , we have @xmath309 , i.e. , the proper time of the observer , on board of the spaceship , traveling in the center of the second bubble , is equal to the time coordinate , @xmath310 .
the spaceship will arrive at @xmath272 in the temporal and spatial intervals given by @xmath311 and @xmath312 , respectively . as in the analysis of the first bubble
, the separation between the departure , at @xmath273 , and the arrival @xmath272 , will be spatial if the analogous relationship of eq.([spatialcharacter ] ) is verified .
therefore , the temporal order between arrival and departure is also not well - defined .
as will be verified below , when @xmath2 and @xmath313 decrease and @xmath310 increases , @xmath5 will decrease and a spaceship will arrive at @xmath272 at @xmath314 . in fact , one may prove that it may arrive at @xmath315 . since the objective is to verify the appearance of ctcs , in principle ,
one may proceed with some approximations . for simplicity , consider that @xmath103 and @xmath308 , and consequently @xmath269 and @xmath316 are enormous , so that @xmath317 and @xmath318 . in this limit
, we obtain the approximation @xmath319 , i.e. , the journey of the first bubble from @xmath272 to @xmath273 is approximately instantaneous
. consequently , taking into account the lorentz transformation , we have @xmath320 and @xmath321 . to determine @xmath322 , which corresponds to the second bubble at @xmath272 , consider the following considerations :
since the acceleration is enormous , we have @xmath323 and @xmath324 , therefore @xmath325 and @xmath326 , from which one concludes that @xmath327 krasnikov discovered an interesting feature of the warp drive , in which an observer in the center of the bubble is causally separated from the front edge of the bubble .
therefore he / she can not control the alcubierre bubble on demand .
krasnikov proposed a two - dimensional metric @xcite , which was later extended to a four - dimensional model @xcite .
one krasnikov tube in two dimensions does not generate ctcs .
but the situation is quite different in the 4-dimensional generalization , which we present for self - consistency and self - completeness .
soon after the krasnikov two - dimensional solution , everett and roman @xcite generalized the analysis to four dimensions , denoting the solution as the _ krasnikov tube_. consider that the 4-dimensional modification of the metric begins along the path of the spaceship , which is moving along the @xmath328-axis , occurring at position @xmath328 at time @xmath329 , the time of passage of the spaceship .
also assume that the disturbance in the metric propagates radially outward from the @xmath328-axis , so that causality guarantees that at time @xmath5 the region in which the metric has been modified can not extend beyond @xmath330 , where @xmath331 .
the modification in the metric should also not extend beyond some maximum radial distance @xmath332 from the @xmath328-axis .
thus , the metric in the 4-dimensional spacetime , written in cylindrical coordinates , is given by @xcite @xmath333 with @xmath334 \ , .
\label{4d : form}\ ] ] for @xmath335 one has a tube of radius @xmath336 centered on the @xmath328-axis , within which the metric has been modified .
this structure is denoted by the _
krasnikov tube_. in contrast with the alcubierre spacetime metric , the metric of the krasnikov tube is static once it has been created .
the stress - energy tensor element @xmath337 given by @xmath338 \,,\ ] ] can be shown to be the energy density measured by a static observer @xcite , and violates the wec in a certain range of @xmath83 , i.e. , @xmath339 . to verify the violation of the wec , consider the energy density in the middle of the tube and at a time long after it s formation , i.e. , @xmath340 and @xmath341 , respectively . in this region
we have @xmath342 , @xmath343 and @xmath344 . with this simplification
the form function , eq . ( [ 4d : form ] ) , reduces to @xmath345 consider the following specific form for @xmath346 @xcite given by @xmath347 + 1 \right \ } \,,\ ] ] so that the form function of eq .
( [ 4d : midtube - form ] ) is provided by @xmath348 + 1 \right \ } \,.\ ] ] choosing the following values for the parameters : @xmath349 , @xmath350 and @xmath351 , it can be shown that the negative character of the energy density is manifest in the immediate inner vicinity of the tube wall .
now , using two such tubes it is a simple matter , in principle , to generate ctcs .
the analysis is similar to that of the warp drive , so that it will be treated in summary .
imagine a spaceship traveling along the @xmath328-axis , departing from a star , @xmath272 , at @xmath173 , and arriving at a distant star , @xmath273 , at @xmath274 .
an observer on board of the spaceship constructs a krasnikov tube along the trajectory .
it is possible for the observer to return to @xmath272 , traveling along a parallel line to the @xmath328-axis , situated at a distance @xmath87 , so that @xmath352 , in the exterior of the first tube . on the return trip
, the observer constructs a second tube , analogous to the first , but in the opposite direction , i.e. , the metric of the second tube is obtained substituting @xmath328 and @xmath5 , for @xmath353 and @xmath354 , respectively in eq .
( [ 4d - krasnikov - metric ] ) .
the fundamental point to note is that in three spatial dimensions it is possible to construct a system of two non - overlapping tube separated by a distance @xmath87 .
after the construction of the system , an observer may initiate a journey , departing from @xmath272 , at @xmath306 and @xmath355 .
one is only interested in the appearance of ctcs in principle , therefore the following simplifications are imposed : @xmath356 and @xmath357 are infinitesimal , and the time to travel between the tubes is negligible . for simplicity , consider the velocity of propagation close to that of light speed
. using the second tube , arriving at @xmath273 at @xmath358 and @xmath274 , then travelling through the first tube , the observer arrives at @xmath272 at @xmath173 .
the spaceship has completed a ctc , arriving at @xmath272 before it s departure .
gtr has been an extremely successful theory , with a well established experimental footing , at least for weak gravitational fields .
it s predictions range from the existence of black holes , gravitational radiation to the cosmological models , predicting a primordial beginning , namely the big - bang .
however , it was seen that it is possible to find solutions to the efes , with certain ease , which generate ctcs .
this implies that if we consider gtr valid , we need to include the _ possibility _ of time travel in the form of ctcs .
a typical reaction is to exclude time travel due to the associated paradoxes .
but the paradoxes do not prove that time travel is mathematically or physically impossible .
consistent mathematical solutions to the efes have been found , based on plausible physical processes .
what they do seem to indicate is that local information in spacetimes containing ctcs are restricted in unfamiliar ways . the grandfather paradox , without doubt ,
does indicate some strange aspects of spacetimes that contain ctcs .
it is logically inconsistent that the time traveler murders his grandfather .
but , one can ask , what exactly impeded him from accomplishing his murderous act if he had ample opportunities and the free - will to do so .
it seems that certain conditions in local events are to be fulfilled , for the solution to be globally self - consistent .
these conditions are denominated _ consistency constraints _ @xcite . to eliminate the problem of free - will , mechanical systems were developed as not to convey the associated philosophical speculations on free - will @xcite .
much has been written on two possible remedies to the paradoxes , namely the principle of self - consistency @xcite and the chronology protection conjecture @xcite .
one current of thought , led by igor novikov , is the principle of self - consistency , which stipulates that events on a ctc are self - consistent , i.e. , events influence one another along the curve in a cyclic and self - consistent way . in the presence of ctcs the distinction between past and future events
are ambiguous , and the definitions considered in the causal structure of well - behaved spacetimes break down .
what is important to note is that events in the future can influence , but can not change , events in the past .
the principle of self - consistency permits one to construct local solutions of the laws of physics , only if these can be prolonged to a unique global solution , defined throughout non - singular regions of spacetime .
therefore , according to this principle , the only solutions of the laws of physics that are allowed locally , reinforced by the consistency constraints , are those which are globally self - consistent .
hawking s chronology protection conjecture @xcite is a more conservative way of dealing with the paradoxes .
hawking notes the strong experimental evidence in favor of the conjecture from the fact that `` we have not been invaded by hordes of tourists from the future '' .
an analysis reveals that the value of the renormalized expectation quantum stress - energy tensor diverges in the imminence of the formation of ctcs .
this conjecture permits the existence of traversable wormholes , but prohibits the appearance of ctcs .
the transformation of a wormhole into a time machine results in enormous effects of the vacuum polarization , which destroys it s internal structure before attaining the planck scale .
nevertheless , li has shown given an example of a spacetime containing a time machine that might be stable against vacuum fluctuations of matter fields @xcite , implying that hawking s suggestion that the vacuum fluctuations of quantum fields acting as a chronology protection might break down .
there is no convincing demonstration of the chronology protection conjecture , but the hope exists that a future theory of quantum gravity may prohibit ctcs .
visser still considers the possibility of two other conjectures @xcite .
the first is the radical reformulation of physics conjecture , in which one abandons the causal structure of the laws of physics and allows , without restriction , time travel , reformulating physics from the ground up .
the second is the boring physics conjecture , in which one simply ceases to consider the solutions to the efes generating ctcs .
perhaps an eventual quantum gravity theory will provide us with the answers .
but , as stated by thorne @xcite , it is by extending the theory to it s extreme predictions that one can get important insights to it s limitations , and probably ways to overcome them .
therefore , time travel in the form of ctcs , is more than a justification for theoretical speculation , it is a conceptual tool and an epistemological instrument to probe the deepest levels of gtr and extract clarifying views .
fsnl acknowledges partial financial support of the fundao para a cincia e tecnologia through the grants ptdc / fis/102742/2008 and cern / fp/109381/2009 .
f. lobo and p. crawford , `` time , closed timelike curves and causality , '' in the nature of time : geometry , physics and perception , nato science series ii .
mathematics , physics and chemistry - vol . * 95 * , kluwer academic publishers , r. buccheri et al .
eds , pp.289 - 296 ( 2003 ) [ arxiv : gr - qc/0206078 ] .
f. lobo and p. crawford , `` weak energy condition violation and superluminal travel , '' current trends in relativistic astrophysics , theoretical , numerical , observational , lecture notes in physics * 617 * , springer - verlag publishers , l. fernndez et al .
eds , pp . 277291 ( 2003 ) [ arxiv : gr - qc/0204038 ] .
j. p. s. lemos , f. s. n. lobo and s. q. de oliveira , phys .
d * 68 * , 064004 ( 2003 ) .
f. j. tipler , phys .
lett * 37 * , 879 - 882 ( 1976 ) .
g. dotti , j. oliva , and r. troncoso , phys .
d * 75 * , 024002 ( 2007 ) ; h. maeda and m. nozawa , phys .
d * 78 * , 024005 ( 2008 ) .
l. a. anchordoqui and s. e. p bergliaffa , phys .
d * 62 * , 067502 ( 2000 ) ; k. a. bronnikov and s .- w .
kim , phys .
d * 67 * , 064027 ( 2003 ) ; m. la camera , phys .
b573 * , 27 - 32 ( 2003 ) ; f. s. n. lobo , phys . rev . *
d75 * , 064027 ( 2007 ) .
k. k. nandi , b. bhattacharjee , s. m. k. alam and j. evans , phys .
d * 57 * , 823 ( 1998 ) ; l. a. anchordoqui , s. e. perez bergliaffa and d. f. torres , phys .
d * 55 * , 5226 ( 1997 ) ; a. g. agnese and m. la camera , phys .
d * 51 * , 2011 ( 1995 ) ; k. k. nandi , a. islam and j. evans , phys .
d * 55 * , 2497 ( 1997 ) ; a. bhattacharya , i. nigmatzyanov , r. izmailov and k. k. nandi , class .
* 26 * , 235017 ( 2009 ) ; f. s. n. lobo and m. a. oliveira , phys .
d * 81 * , 067501 ( 2010 ) .
f. s. n. lobo and m. a. oliveira , phys .
d * 80 * , 104012 ( 2009 ) .
r. garattini and f. s. n. lobo , class .
* 24 * , 2401 ( 2007 ) ; r. garattini and f. s. n. lobo , phys .
b * 671 * , 146 ( 2009 ) . c. g. boehmer , t. harko and f. s. n. lobo , phys .
d * 76 * , 084014 ( 2007 ) ; c. g. boehmer , t. harko and f. s. n. lobo , class .
grav . * 25 * , 075016 ( 2008 ) .
f. s. n. lobo and p. crawford , class .
. grav .
* 21 * , 391 ( 2004 ) ; f. s. n. lobo and p. crawford , class .
* 22 * , 4869 ( 2005 ) ; j. p.
s. lemos and f. s. n. lobo , phys .
d * 69 * , 104007 ( 2004 ) ; f. s. n. lobo , gen .
* 37 * , 2023 ( 2005 ) ; j. p. s. lemos and f. s. n. lobo , phys .
d * 78 * , 044030 ( 2008 ) .
s. sushkov , phys .
d * 71 * , 043520 ( 2005 ) ; f. s. n. lobo , phys .
rev . * d71 * , 084011 ( 2005 ) ; f. s. n. lobo , phys .
rev . * d71 * , 124022 ( 2005 ) ; j. a. gonzalez , f. s. guzman , n. montelongo - garcia and t. zannias , phys .
d * 79 * , 064027 ( 2009 ) ; a. debenedictis , r. garattini and f. s. n. lobo , phys .
d * 78 * , 104003 ( 2008 ) ; f. s. n. lobo , phys .
rev . * d73 * , 064028 ( 2006 ) ; n. m. garcia and t. zannias , phys .
d * 78 * , 064003 ( 2008 ) ; f. s. n. lobo , phys .
d * 75 * , 024023 ( 2007 ) .
n. montelongo garcia and t. zannias , class .
* 26 * , 105011 ( 2009 ) .
f. s. n. lobo , class .
grav . * 25 * , 175006 ( 2008 ) .
t. harko , z. kovacs and f. s. n. lobo , phys .
d * 78 * , 084005 ( 2008 ) ; t. harko , z. kovacs and f. s. n. lobo , phys .
d * 79 * , 064001 ( 2009 ) .
j. p. s. lemos ,
f. s. n. lobo and s. quinet de oliveira , phys .
d * 68 * , 064004 ( 2003 ) .
f. s. n. lobo , `` exotic solutions in general relativity : traversable wormholes and warp drive spacetimes , '' arxiv:0710.4474 [ gr - qc ]
. m. s. morris , k. s. thorne and u. yurtsever , `` wormholes , time machines and the weak energy condition , '' phy . rev . lett . * 61 * , 1446 ( 1988 ) . j. l. friedman , m. s. morris , i. d. novikov , f. echeverria , g. klinkhammer , k. s. thorne , and u. yurtsever , phys .
d * 42 * , 1915 - 1930 ( 1990 ) ; m. visser , in the future of theoretical physics and cosmology , cambridge university press , edited by g. gibbons _
et al _ , pp.161 - 176 ( 2003 ) . m. visser , phys . rev .
d * 41 * 1116 ( 1990 ) .
k. d. olum , phys .
d * 61 * 124022 ( 2002 ) .
a. ori , phys .
lett . * 95 * , 021101 ( 2005 ) .
m. alcubierre , class .
. grav . * 11 * , l73-l77 ( 1994 ) . j. natrio , class .
. grav . * 19 * , 1157 , ( 2002 ) ; f. s. n. lobo and m. visser , class .
* 21 * , 5871 ( 2004 ) ; d. h. coule , class .
15 * , 2523 - 2527 ( 1998 ) ; w. a. hiscock , class .
* 14 * , l183 ( 1997 ) ; c. clark , w. a. hiscock and s. l. larson , class .
* 16 * , 3965 ( 1999 ) ; p. f. gonzlez - daz , phys .
d * 62 * , 044005 ( 2000 ) ; c. van den broeck , class .
* 16 * , 3973 , ( 1999 ) .
a. e. everett , phys .
d * 53 * 7365 ( 1996 ) .
s. v. krasnikov , phys .
d * 57 * , 4760 ( 1998 ) .
a. e. everett and t. a. roman , phys .
d * 56 * , 2100 ( 1997 ) .
j. earman , _ bangs , crunches , whimpers , and shrieks : singularities and acausalities in relativistic spacetimes _ , oxford university press ( 1995 ) .
a. carlini , v.p .
frolov , m. b. mensky , i. d. novikov and h. h. soleng , int . j. mod .
d * 4 * 557 ( 1995 ) ; erratum - ibid d*5 * 99 ( 1996 ) .
a. carlini and i. d. novikov , int . j. mod .
d * 5 * 445 ( 1996 ) .
li , phys .
d * 50 * , r6037 ( 1994 ) .
k. s. thorne , in _ general relativity and gravitation _ , proceedings of the 13th conference on general relativity and gravitation , edited by r. j. gleiser et al ( institute of physics publishing , bristol , 1993 ) , p. 295 . | the conceptual definition and understanding of time , both quantitatively and qualitatively is of the utmost difficulty and importance . as time
is incorporated into the proper structure of the fabric of spacetime , it is interesting to note that general relativity is contaminated with non - trivial geometries which generate _ closed timelike curves_. a closed timelike curve ( ctc ) allows time travel , in the sense that an observer that travels on a trajectory in spacetime along this curve , may return to an event before his departure .
this fact apparently violates causality , therefore time travel and it s associated paradoxes have to be treated with great caution .
the paradoxes fall into two broad groups , namely the _ consistency paradoxes _ and the _ causal loops_. a great variety of solutions to the einstein field equations containing ctcs exist and it seems that two particularly notorious features stand out .
solutions with a tipping over of the light cones due to a rotation about a cylindrically symmetric axis and solutions that violate the energy conditions .
all these aspects are analyzed in this review paper . |
in recent years , there has been an increasing interest in cavity optomechanical systems for a wide range of both experimental and theoretical [email protected] importance of optomechanical systems is due to their potential applications in different topics of physics.they are promising candidates for studying quantum effects in the mesoscopic and macroscopic scales@xcite , detection and interferometry of gravitational waves @xciteand measurement of small displacements@xcite . in a cavity
optomechanical system the radiation pressure exerted by the electromagnetic field induces a coupling between the intensity of the cavity field and the mechanical motion of a movable mirror .
it is pointed out that the necessary condition for observing quantum phenomena for the oscillating mirror , as a mesoscopic or macroscopic object , is the preparation of the mechanical oscillator at low phonon occupancy .
although the ground state cooling of the mirror has not yet been achieved experimentally , it has been shown theoretically@xcite that the ground state cooling of the mirror is possible in the resolved sideband regime where the cavity bandwith is less than mechanical oscillation frequency of the mirror .
another quantum phenomenon expected to be observed in optomechanical sytems is the mixing between the fluctuations of the cavity field around the steady state and the mechanical mode which leads to the normal mode splitting(nms ) in the displacement spectrum of the mirror@xcite .
it is well known from cavity qed that nms is an absolute evidence of strong coupling between the subsytems with energy exchange taking place on a time scale faster than the decoherence of each mode .
the appearance of nms in optomechanical systems is known as a direct consequence of ground state cooling of the mirror@xcite .
one of the most important characteristics of optomechanical systems is their nonlinear optical peroperties .
the strong interaction between the cavity filed and the mechanical oscillations makes the cavity behave like a nonlinear medium since the length of the cavity depends upon the intensity of the field in analogous way to the optical length of a nonlinear material@xcite .
it has been shown@xcite that within the born - openheimer approximation the back action of the oscillating mirror in one edge of an optical fabry - perot cavity , driven by a nearly - resonant laser field , induces an intracavity kerr - like nonlinearity . besides this intrinsic nonlinearity ,
the presence of an optical parametric amplifier ( opa ) @xcite or the optical kerr medium @xcite inside the cavity has opened up a new domain for combining nonlinear optics and optomechanics towards the enhancement of quantum effects .
it has been demonstrated@xcite that the presence of an opa in the cavity causes a strong coupling between the oscillating mirror and the cavity mode resulting from increasing the intracavity photon number .
thus the enhancement of radiation pressure - induced coupling not only considerably improves the cooling of the mechanical oscillator but also makes the observation of the nms of the movable mirror and the output field mode accesible@xcite . on the other hand , when the optomechanical cavity contains an optical kerr medium with @xmath0 nonlinearity , due to the photon - photon repulsion and the reduction of the cavity photon fluctuations , the normal mode splitting weakens and the effective temperature of the moving mirror increases@xcite .
the above mentiond interesting results motivated us towards investigation of an optomechanical system which contains a nonlinear crystal consisting of a kerr medium and a degenerate opa .
we will show that in the system under consideration the competition between the increasing of intracavity photon number due to the parametric gain process and the reduction of the number of photons due to the photon blockade mechanism arising from the presence of the kerr medium leads to some new interesting effects in the dynamics of the movable mirror in the low photon number limit and the resolved sideband regime .
the paper is structured as follows . in sec.[sec1 ] we describe the model , derive the quantum langevin equations of motion for the system operators and find their steady - state mean values . in sec.[sec2 ] we linearize the quntum langevin equations of motion around the steady - state mean values , and analyze the stability conditions of the system . in sec.[sec3 ] we calculate the spectrum of small fluctuations in the position of the oscillating mirror , the effective frequency and effective damping rate of the mechanical oscilltor , and analyze the influence of the nonlinear gain and anharmonicity parameter on them . in sec.[field ] we investigate the influences of the parametric gain and kerr nonlinearity on the intensity and quadrature squeezing of the transmitted field . in sec.[ent ] we examine the entanglement between the optical and the mechanical mode .
finally , we summarize our conclusions in sec.[sum ] .
we consider a kerr - down conversion optomechanical system composed of a degenerate opa and a nonlinear kerr medium placed within a fabry - perot cavity formed by a fixed partially transmitting mirror and one movable perfectly reflecting mirror in equilibrium with a thermal bath at temperature @xmath1 .
the movable mirror is free to move along the cavity axis and is treated as a quantum mechanical harmonic oscillator with effective mass @xmath2 , frequency @xmath3 and energy decay rate @xmath4(where @xmath5 is the mechanical quality factor).the cavity field is coherently driven by an input laser field with frequency @xmath6 and amplitude @xmath7 through the fixed mirror .
furthermore , the system is pumped by a coupling field to produce parametric oscillation and induce the kerr nonlinearity in the cavity . in our investigation
, we restrict the model to the case of single - cavity and mechanical modes@xcite .
the single cavity - mode assumption is justified in the adiabatic limit , i.e. , @xmath8 which @xmath9 is the speed of light in vacuum and @xmath10 is the cavity length in the absence of the cavity field .
we also assume that the induced resonance frequency shift of the cavity and the nonlinear parameter of the kerr medium are much smaller than the longitudinal - mode spacing in the cavity .
furthermore , one can restrict to a single mechanical mode when the detection bandwidth is chosen such that it includes only a single , isolated , mechanical resonance and mode - mode coupling is negligible .
it should be noted that in the adiabatic limit , the number of photons generated by the casimir , retardation , and doppler effects is negligible @xcite . under these conditions , the total hamiltonian of the system in a frame rotating at the laser frquency @xmath6can be written as @xmath11 where , [ hamil ] @xmath12 the first two terms in @xmath13 are , respectively , the free hamiltonian of the cavity field with annihilation(creation ) operator @xmath14 and the movable mirror with resonance frequency @xmath3 and effective mass @xmath2 , the third term describes the optomechanical coupling between the cavity field and the mechanical oscillator due to radiation pressure force , and the last term in @xmath13 describes the driving of the intracavity mode with the input laser .
also , the two terms in @xmath15 describes , respectively , the coupling of the intracavity field with the opa and the kerr medium ; @xmath16 is the nonlinear gain of the opa which is proportional to the pump power driving amplitude , @xmath17 is the phase of the field driving the opa , and @xmath18 is the anharmonicity parameter proportional to the third order nonlinear susceptibility @xmath19 of the kerr medium .
the input laser field populates the intracavity mode through the partially transmitting mirror , then the photons in the cavity will exert a radiation pressure force on the movable mirror . in a realistic treatment of the dynamics of the system ,
the cavity - field damping due to the photon - leakage through the incomplete mirror and the brownian noise associated with the coupling of the oscillating mirror to its thermal bath should be considered . using the input - output formalism of quantum optics@xcite
, we can consider the effects of these sources of noise and dissipation in the quantum langevin equations of motion .
for the given hamiltonian ( [ h ] ) , we obtain the following nonlinear equations of motion [ langevin ] @xmath20 where @xmath21 is the cavity decay rate through the input mirror and @xmath22 is the input vacuum noise operator that is charactrized by the following correlation functions @xcite [ noise1 ] @xmath23 the brownian noise operator @xmath24 describes the heating of the mirror by the thermal bath at temperature @xmath25 and is characterized by the following correlation function @xcite @xmath26d\omega,\ ] ] where @xmath27 is the boltzmann constant .
we are interested in the steady - state regime and in small fluctuations with respect to the steady state .
thus we obtain the steady - state mean values of @xmath28 , @xmath29 and @xmath30 as @xmath31 @xmath32 where @xmath33 denotes the new equilibrium position of the movable mirror and @xmath34 is the effective detuning of the cavity which includes both the radiation pressure and the kerr medium effects .
it is obvious that the optical path and hence the cavity detuning are modified in an intensity - dependent way .
the first modification which is a mechanical nonlinearity , arises from the radiation pressure - induced coupling between the movable mirror and the cavity field and the second modification comes from the presence of the nonlinear kerr medium in the optomechanical system .
it has been shown that @xcite for a given sufficiently large pure @xmath19 nonlinearity inside an optical cavity , the effect of photon blockade occurs as a consequence of large phase shifts of the cavity detuning .
it has also been predicted @xcite that the same photon - photon interaction can occur in the strong coupling regime in the optomechnaical systems . in our treatment
the photon - photon interaction due to the radiation pressue is ignorable in comparison with photon - photon interaction due to the kerr nonlinearity . since @xmath35 satisfies a fifth - order equation , it can have five real solutions and hence the system may exhibit multistability for a certain range of parameters .
in order to investigate the dynamics of the system , we need to study the dynamics of small fluctuations near the steady state .
we assume that the nonlinearity in the system is weak and decompose each operator in eq.([langevin ] ) as the sum of its steady - state value and a small fluctuation with zero mean value , @xmath36 inserting the above linearized forms of the system operators into eq.([langevin ] ) , the linearized quantum langevin equations for the fluctuation operators take the form [ fluc ] @xmath37 defining the cavity field quadratures @xmath38 and @xmath39 and the input noise quadratures @xmath40 and @xmath41 eq.([fluc ] ) can be written in the compact matrix form @xmath42 where @xmath43 is the vector of fluctuations , @xmath44 is the vector of noise sources and the matrix @xmath45 is given by @xmath46 where we have defined @xmath47 here , we concentrate on the stationary properties of the system . for this purpose , we should consider the steady - state condition governed by eq.([matrixform ] ) . the system is stable only if the real part of all eigenvalues of the matrix @xmath45 are negative , which is also the requirement of the validity of the linearized method .
the parameter region in which the system is stable can be obtained from the routh - hurwitz criterion@xcite , which gives the following three independent conditions : @xmath48\times\omega_{m}^{2 } ( \gamma_{m } + 2\kappa)+ [ 2\kappa r_{1}\nonumber\\ & + & \omega_{m}^{2}(\gamma_{m } + 4\kappa^{2 } ) ] > 0,\end{aligned}\ ] ] [ routh1 ] where we have defined @xmath49 and @xmath50 as @xmath51
since we are interested in the spectrum of fluctuations in position of the oscillating mirror in the presence of kerr - down conversion nonlinearity , it is convenient to solve the time - domain equation of motion ( [ matrixform ] ) by fourier transforming it into the frequency domain .
the fourier transform of the time - domain operator @xmath52 is @xmath53 then solving the linearized quantum langevin equation for the position fluctuations of the oscillating mirror yields @xmath54 in which @xmath55,\nonumber\\ f_{3}(\omega)&=&f_{2}(-\omega)^{*},\end{aligned}\ ] ] with @xmath56 @xmath57,\nonumber\\ \beta_{2}&=&-4g^{2}+\delta_{1}^{2}-4\eta^{2}|a_{s}|^{4}+ 4i g \eta ( a_{s}^{2}e^{-i\theta}-c.c.).\nonumber\\ \end{aligned}\ ] ] the first term in eq.([deltaq ] ) originates from the thermal noise while the second and third terms arise from the radiation pressure force .
the displacement spectrum of the mirror is defined by @xmath58 using the following correlation functions in the frequency domain @xmath59 @xmath60\delta(\omega+\omega),\ ] ] the displacement spectrum of the oscillating mirror is obtained as @xmath61 where @xmath62 . in order to determine the structure of the displacement spectrum of the moving mirror we need to determine the eigenvalues of the matrix @xmath63 as the solution of eq.([matrixform ] ) in the frequency domain , or the zeros of @xmath64 in the denominator of coefficients in eq.([coefficient ] ) .
first we assume that the radiation pressure coupling @xmath65 is zero and find the eigenvalues of @xmath63 as @xmath66 thus in the absence of optomechanical coupling the effective frequency of the normal modes depends on @xmath67 which is a function of nonlinear coefficiens @xmath18 , @xmath16 and intracavity intensity .
now we assume that the optomechanical coupling @xmath65 is not zero , but the decay rates @xmath68 and @xmath21 are negligibly small in comparison with @xmath3 and @xmath67
. then the zeros of @xmath64 are approximately given by @xmath69 hence , the kerr - down conversion medium can alter the radiation pressure contribution to the nms in two ways .
firstly , both nonlinearity parameters @xmath16 and @xmath18 change the mean intracavity photon number considerably .
secondly the parameter @xmath70 is a function of the gain and anharmonicity parameters . in fig.[fig :
beta ] , by using the experimental parameters from ref.@xcite , we have plotted the mean intracavity photon number @xmath71 and the normalized parameter @xmath72 versus the normalized gain @xmath73 for two different values of anharmonicity parameter @xmath18 , where @xmath74 is the contribution of radiation pressure to the normal modes of displacement spectrum in the bare cavity .
the figure shows that while there is a reduction in the number of intracavity photons ( hence the radiation pressure ) due to the photon - photon repulsion mechanism ( in the presence of kerr medium ) , the contribution of radiation pressure coupling to the normal modes which is contained in @xmath70 can be increased by choosing proper values of initial detuning and nonlinear coefficients @xmath16 , @xmath18 and @xmath17 in such a way that @xmath75 . versus the normalized gain @xmath73 for @xmath76hz ( red solid line ) and @xmath77
hz ( blue dashed line).the parameters are @xmath78 mhz,@xmath79 mm , @xmath80ng , @xmath81 hz , input laser power @xmath82 mw at @xmath83 nm , @xmath84 , and @xmath85 .
, title="fig:",width=3 ] versus the normalized gain @xmath73 for @xmath76hz ( red solid line ) and @xmath77 hz ( blue dashed line).the parameters are @xmath78 mhz,@xmath79 mm , @xmath80ng , @xmath81 hz , input laser power @xmath82 mw at @xmath83 nm , @xmath84 , and @xmath85 . , title="fig:",width=3 ] for bare cavity @xmath86(blue thin curve ) , for cavity with kerr nonlinearity @xmath87 ( orange dashed - dotted curve ) , with gain nonlinearity @xmath88 ( green dashed curve ) and with both nonliearity@xmath89 ( red thick curve ) .
the parameters are @xmath78 mhz,@xmath90 mm , @xmath91ng , @xmath81 hz , input laser power @xmath92 mw at @xmath83 nm , @xmath93 , and @xmath85 . ,
width=336 ] it is well known that the nms in a bare - damped optomechanical system occurs only if @xmath94 ( where @xmath95 ) due to the finite width of the peaks@xcite . in fig.([fig : figc ] ) we have plotted the displacement spectrum of the movable mirror below this threshold ( @xmath94 ) . as expected , the radiation pressure - induced coupling between the oscillating mirror and the fluctuations of the cavity mode does not lead to nms .
the presence of the kerr medium alone dose not change this situation , while in the presence of the gain medium solely , mode splitting occurs only when the parametric gain @xmath16 is large enough . however , as shown in fig.([fig : figc ] ) , the combination of these two nonlinearities leads to the appearance of nms in the displacement spectrum . we can come to this conclusion from the fact that in the peresence of the two nonlinearities , there is a competition between the two quantum processes
; the nonlinear gain medium increases the intracavity photon number and splits up the pump photons into two degenerate subharmonic photons , while the kerr medium shifts the energy levels of the cavity proportional to the anharmonicity parameter @xmath18 which blocks the entrance of a large number of photons into the cavity .
the combination of the energy shift resulting from the kerr medium and increasing the photon number due to the gain medium leads to the appearence of nms .
it is notable that the presence of the gain medium decreases the degree of photon - photon repulsion due to the kerr nonlinearity , and hence increases the parameter @xmath70 in eq.([normal])and the radiation pressure effect on the oscillating mirror . in fig.([fig : dis - g ] ) and fig.([fig : dis - eta ] ) we have plotted the displacement spectrum of the movable mirror above the threshold of splitting ( @xmath94 ) for different values of the gain coefficient @xmath16 and different values of anharmonicity parameter @xmath18 , respectively . as expected , the presence of the gain medium inside the cavity strengthens the coupling between the oscillating mirror and the cavity mode , while the presence of the kerr medium inside the cavity weakens this coupling due to the photon - photon repulsion and shifts the modes far from each other . for different values of the parametric gain :
@xmath96(green thin curve ) , @xmath97hz(blue dashed curve ) , and @xmath98hz(red thick curve ) .
the parameters are
@xmath76 hz , @xmath99 , @xmath85 , @xmath78 mhz,@xmath79 mm , effective mass @xmath80ng , @xmath81 hz , @xmath100 mw , @xmath83 nm , and @xmath101.,width=336 ] for different values of the anharmonicity : @xmath76hz(red thick curve ) @xmath102hz(blue dashed curve ) , and @xmath103hz(green thin curve).the parameters are @xmath104hz , @xmath99 , @xmath85 , @xmath78 mhz,@xmath79 mm , @xmath80ng , @xmath81 hz , @xmath100 mw , @xmath83 nm , and @xmath101.,width=288 ] the radiation - pressure force can modify the dynamics of the mechanical oscillator and therefore the susceptibility of the oscillating mirror can be considered as the susceptibility of an oscillator , with an effective frequency and an effective damping.these quantities contain important information about the quantum behavior of the system .
the radiation presure - induced change in the frequency of the mechanical mode is known as the optical spring effect @xcite , and the radiation presure - induced change in the damping rate of the mirror shows how the optical field acts effectively as a viscous fluid that damps the mirror oscillation and cools its center - of mass motion@xcite . in this section
we investigate the effects of the kerr and the gain nonlinearities on the effective frequency @xmath105 and the effective damping rate @xmath106 of the oscillating mirror .
we can find the mechanical susceptibility of the mirror according to the dependence of the @xmath107 on the fluctuations in the total force on the mirror @xmath108@xcite , @xmath109 according to eq.([deltaq ] ) , @xmath108 consists of a radiation pressure and a brownian motion term .
thus the mechanical susceptibility of the mirror @xmath110 , its effective resonance frequency @xmath111 , and its effective damping rate @xmath112 are , respectively , given by @xmath113 @xmath114 @xmath115 in figs.[fig : oeff](a ) and [ fig : oeff](b ) we have plotted the normalized effective frequency of the oscillating mirror @xmath116 versus the normalized frequency @xmath117 for @xmath99 and for different values of @xmath16 [ fig.([fig : oeff]a ) ] and different values of @xmath18 [ fig.([fig : oeff]b ) ] .
as is seen , although the gain medium does not change the effective frequency of the moving mirror considerably , the energy shift of the cavity mode due to the kerr nonlinearity manifests itself in the response frequency of the oscillating mirror .
this means that the effective coupling between the cavity and the mechanical modes ( and therefore the cooling of the mirror ) takes place when the effective detuning of the cavity @xmath118 ( but not @xmath119 ) is chosen near the oscillating mirror frequency .
[ ht ] versus the @xmath117 ( a ) for @xmath76hz and for different values of the gain parameter : @xmath96 ( blue dashed curve ) , and @xmath104hz ( red solid curve ) , ( b ) for @xmath104hz and for different values of anharmonicity parameter : @xmath120 ( blue dashed curve ) , and @xmath76hz ( red solid curve ) .
the parameters are @xmath78 mhz , @xmath121 mm , @xmath80 ng , @xmath81 hz , @xmath82 mw , @xmath101 , @xmath99 , and @xmath85 . , title="fig:",width=3 ] versus the @xmath117 ( a ) for @xmath76hz and for different values of the gain parameter : @xmath96 ( blue dashed curve ) , and @xmath104hz ( red solid curve ) , ( b ) for @xmath104hz and for different values of anharmonicity parameter : @xmath120 ( blue dashed curve ) , and @xmath76hz ( red solid curve ) .
the parameters are @xmath78 mhz , @xmath121 mm , @xmath80 ng , @xmath81 hz , @xmath82 mw , @xmath101 , @xmath99 , and @xmath85 . ,
title="fig:",width=3 ] [ ht ] versus the @xmath117 ( a ) for @xmath76hz and for different values of gain parameter : @xmath96 ( blue dashed curve ) , and @xmath104hz ( red solid curve ) , ( b ) for @xmath104hz and for different values of anharmonicity parameter : @xmath120 ( blue dashed curve ) , and @xmath76hz ( red solid curve ) .
the parameters are @xmath78 mhz , @xmath121 mm , @xmath80 ng , @xmath81 hz , @xmath82 mw , @xmath101 , @xmath99 , and @xmath85 . , title="fig:",width=3 ] versus the @xmath117 ( a ) for @xmath76hz and for different values of gain parameter : @xmath96 ( blue dashed curve ) , and @xmath104hz ( red solid curve ) , ( b ) for @xmath104hz and for different values of anharmonicity parameter : @xmath120 ( blue dashed curve ) , and @xmath76hz ( red solid curve ) .
the parameters are @xmath78 mhz , @xmath121 mm , @xmath80 ng , @xmath81 hz , @xmath82 mw , @xmath101 , @xmath99 , and @xmath85 . ,
title="fig:",width=3 ] in figs.[fig : dmpeff](a ) and [ fig : dmpeff](b ) we have plotted the normalized effective damping rate of the oscillating mirror @xmath122 versus the normalized frequency @xmath117 for @xmath99 and for different values of @xmath16 [ fig.[fig : dmpeff](a ) ] and different values of @xmath18 [ fig.[fig : dmpeff](b ) ] . as is seen , by increasing the gain parameter the effective damping rate of the mirror motion increases while increasing the anharmonicity parameter not only leads to the frequency shift but also decreases the effective damping rate . here
we examine the effects of the kerr and the gain nonlinearities on the back - action ground - state cooling of the oscillating mirror . in order to investigate the cooling of the mirror it is sufficient to consider the mean number of quanta of the vibrational excitation of the mirror as defined by@xcite @xmath123 where we expand the effective frequency and effective damping rate of the oscillating mirror in a taylor series around @xmath124 and keep only the leading terms in the respective expansions @xmath125 the ground - state cooling
is approached if @xmath126 . in figs.[fig : n , teff](a ) and [ fig : n , teff](b ) we have plotted , respectively , @xmath127 and @xmath128 for the initial equilibrium temperature @xmath129k versus @xmath130 and for different values of @xmath16 and @xmath18 .
it is evident that as @xmath16 increases the system cools down to the ground - state ( @xmath131 ) while as @xmath18 increases the effective temperature of the system increases .
the heating of the system in the presence of the kerr medium is due to the energy shift of the cavity which weakens the field - mirror coupling and reduces the number of intracavity photons because of the photon - photon repulsion mechanism .
[ ht ] and ( b ) the effective temperature @xmath128 for @xmath132hz,@xmath77hz ( blue dotted - dashed curve ) , @xmath132hz,@xmath76hz ( red thick curve ) and @xmath133hz ( green thin curve ) .
the parameters are @xmath78 mhz , @xmath134 mm , @xmath80 ng , @xmath81 hz , @xmath82 mw , @xmath135 , and @xmath85 . , title="fig:",width=3 ] and ( b ) the effective temperature @xmath128 for @xmath132hz,@xmath77hz ( blue dotted - dashed curve ) , @xmath132hz,@xmath76hz ( red thick curve ) and @xmath133hz ( green thin curve ) .
the parameters are @xmath78 mhz , @xmath134 mm , @xmath80 ng , @xmath81 hz , @xmath82 mw , @xmath135 , and @xmath85 . , title="fig:",width=3 ]
the intensity spectrum of the transmitted field is given by the fourier transform of the two time correlation functions @xmath136 : @xmath137
solving the matrix equation ( [ matrixform ] ) we obtain the solution for the cavity - field fluctuation operator @xmath138 and then using the input - output relation @xmath139 , we obtain the fluctuations of the output field to be @xmath140 and @xmath141^{\dagger}$ ] , where @xmath142 in eq.([aout ] ) the first term stems from the mechanical oscillator thermal noise , while the other two terms are from the input vacuum noise . making use of the correlation peroperties for the noise forces and eqs.([aout ] ) and ( [ a - coeff ] ) , we obtain the output field spectrum as @xmath143\vert v _ { 1}(-\omega)\vert ^{2}.\ ] ] the quadrature noise spectrum of the transmitted filed
is given by@xcite @xmath144 where @xmath145 is the fourier trnasform of the output quadrature , with @xmath146 as its externally controllable phase angle which is experimentally measurable in a homodyne detection [email protected] the system under consideration quadrature noise spectrum is given by @xmath147 where @xmath148 v_{1}(\omega)v_{1}(-\omega))\nonumber\\ & + & v_{2}(\omega)v_{3}(-\omega)\rbrace,\\ c_{aa^{\dagger } } ^{out } & = & \lbrace m \hbar \gamma_{m}\omega[1+\coth(\frac{\hbar\omega}{2k_{b}t})]\vert v_{1}(\omega)\vert ^{2}\nonumber\\ & + & \vert v_{2}(\omega)\vert ^{2}\rbrace,\\ c_{a^{\dagger}a}^{out } & = & \lbrace m \hbar \gamma_{m}\omega[1+\coth(\frac{\hbar\omega}{2k_{b}t})]\vert v_{1}(-\omega)\vert ^{2}\nonumber\\ & + & \vert v_{3}(-\omega)\vert ^{2 } \rbrace .
\end{aligned}\ ] ] we define the optimum quadrature squeezing @xmath149 by choosing @xmath150 in such a way that @xmath151 .
this yields @xmath152 then , substituting back into eq.([squeezing ] ) , we obtain the optimized squeezing spectrum as @xmath153 in figs.[fig : fi - g](a ) and [ fig : fi - g](b ) we have plotted , respectively , the intensity and the optimized squeezing spectra of the transmitted field for different values of the gain nonlinearity versus the normalized response frequency @xmath117 .
as can be seen increasing the gain parameter @xmath16 causes the intensity and the degree of squeezing of the transmitted field to increase .
the shift in the intensity spectrum is due to the energy shift of the cavity in the presence of the kerr medium . in figs.[fig : fi - eta](a ) and [ fig : fi - eta](b ) we have plotted , respectively , the intensity and the optimized squeezing spectra of the transmitted field for different values of the anharmonicity parameter @xmath18 versus the normalized response frequency @xmath117 .
it can be seen as @xmath18 increases the intensity and the degree of squeezing of the transmitted field increases .
[ ht ] for @xmath154 hz ( red thick curve ) , @xmath132 hz(blue dashed curve ) , and @xmath155 hz(green thin curve ) .
the parameters are @xmath76hz , @xmath78 mhz , @xmath134 mm , @xmath80 ng , @xmath81 hz , @xmath82 mw , @xmath135 , and @xmath85 . , title="fig:",width=3 ] for @xmath154 hz ( red thick curve ) , @xmath132 hz(blue dashed curve ) , and @xmath155 hz(green thin curve ) . the parameters are @xmath76hz , @xmath78 mhz , @xmath134 mm , @xmath80 ng , @xmath81 hz , @xmath82 mw , @xmath135 , and @xmath85 . , title="fig:",width=3 ] [ ht ] for @xmath76 hz ( red thick curve ) , @xmath102 hz(blue dashed curve ) , and @xmath77 hz(green thin curve ) .
the parameters are @xmath156hz , @xmath78 mhz , @xmath134 mm , @xmath80 ng , @xmath81 hz , @xmath82 mw , @xmath135 , and @xmath85 . ,
title="fig:",width=3 ] for @xmath76 hz ( red thick curve ) , @xmath102 hz(blue dashed curve ) , and @xmath77 hz(green thin curve ) .
the parameters are @xmath156hz , @xmath78 mhz , @xmath134 mm , @xmath80 ng , @xmath81 hz , @xmath82 mw , @xmath135 , and @xmath85 . , title="fig:",width=3 ]
it has been shown previously @xcite that within each optomechanical cavity the photon - phonon entanglement can be generated by means of radiation pressure .
here we are interested in the generation of stationary entanglement between the movable mirror and the cavity field in the presence of kerr - down conversion medium . since our system is in a gaussian state and the linear nature of eq.([matrixform ] )
preserves the gaussian nature of the initial state of the system , we can use the logarithmic negativity measure @xcite to quantify the entanglement .
when the stability conditions of eq.([routh1 ] ) are fulfilled , we can solve eq.([matrixform ] ) for the @xmath157 stationary correlation correlation matrix ( cm ) @xmath158 @xmath159 where the elements of the correlation matrix @xmath158 are defined as @xmath160 and @xmath161 $ ] is the diagonal diffusion matrix in which we used the following apporximation @xmath162 where @xmath163^{-1}$ ] .
then the photon - phonon entanglement can be quantified by the logarithmic negativity @xmath164 as @xmath165,,\ ] ] where @xmath166^{1/2}$ ] , is the lowest symplectic eigenvalue of the partialtranspose of the @xmath167 cm , @xmath158 , with @xmath168 , and we used the @xmath169 block form of the cm @xmath170 with @xmath171 associated to the oscillating mirror , @xmath172 to the cavity mode , and @xmath173 describing the optomechanical correlations . versus the normalized detuning @xmath130 for @xmath174 ( green thick curve ) , @xmath175(red dashed - dotted curve ) , @xmath176(blue dashed curve ) and @xmath177(orange thin curve ) .
the parameters are @xmath78 mhz , @xmath79 mm , @xmath80 ng @xmath81 hz , @xmath178 mw , @xmath179 , and @xmath85.,width=288 ] versus the normalized detuning @xmath130 for @xmath180 hz(green solid curve ) , @xmath181 hz(red dashed - dotted curve ) , and @xmath182hz(blue dashed curve ) .
the parameters are @xmath76 hz , @xmath78 mhz , @xmath79 mm , @xmath80 ng , @xmath81 hz , @xmath178 mw , @xmath179 , and @xmath85 . , width=288 ] in fig.([fig : figent ] )
we have plotted @xmath164 versus the normalized detuning @xmath130 for a bare cavity , a cavity that contains only the kerr medium , a cavity that contains only the gain medium and a cavity that contains both the kerr and gain medium .
we see that , as expected , while the kerr medium reduces the degree of phonon - photon entanglement , the gain medium increases it considerably .
this result arises from the fact that the radiation pressure , which is responsible for the phonon - photon entanglement , increases considerably in the presence of the gain nonlinearity while decreases in the presence of the kerr nonlinearity .
therefore similarly to what happens for the mirror cooling process there is a competition between the gain and kerr nonlinearities for increasing and decreasing quantum effects , respectively . in fig.[fig : figent2 ] we have plotted @xmath164 versus the normalized detuning for different values of the gain nonlinearity .
it can be seen that as @xmath16 increases the degree of entanglement increases .
it should be noted that we can not increase @xmath16 arbitrarily because of the stability condition of the system .
in this paper we considered the canonical optomechanical cavity composed of a partiallly transmitting mirror and one perfectly reflecting movable mirror containing a kerr - down conversion nonlinear medium .
for this system we investigated the influence of the two nonlinearities on the dynamics of the oscillating mirror , the intensity and squeezing spectra of the transmitted field and the steady - state mirror - field entanglement . for the dynamics of the oscillating mirror
we have found that in the peresence of the two nonlinearities , there is a competition between the two quantum processes ; the splitting of the pump photons in two degenerate subharmonic photons as well as increasing the intracavity photon number due to the nonlinear gain medium , and shifting the energy level of the cavity proportional to the anharmonicity parameter @xmath18 which blocks the entrance of a large number of photons into the cavity .
this competition can lead to enhancement of the radiation pressure contribution to the normal modes of the cavity and nms below the threshold @xmath183 .
also , it was demonstrated that in the presence of the kerr - down conversion nonlinearity the ground - state cooling of the mechanical mirror is possible .
although with increasing the anharmonicity parameter @xmath18 the vibrational excitation and effective temperature of the mirror increases , the enhancement of the gain medium can compensate the increasing of the phonon number to some extent . in the investigation of the intensity spectrum and quadrature squeezing of the transmitted field
, it was shown that the range of response frequency of the field undergoes a shift because of anharmonicity parameter @xmath18 but in this range of the frequency both the kerr and gain media increase the degree of squeezing . finally , we investigated the influence of the kerr - down conversion on the degree of stationary entanglement between the cavity and the mechanical modes .
it was shown that the kerr nonlinearity reduces the degree of entanglement considerably , while the gain medium increases it .
the authors wish to thank the office of graduate studies of the university of isfahan for their support .
t. j. kippenberg and k. j. vahala , opt .
express * 15 * , 17172(2007 ) .
s. groblacher , k. hammerer , m. r. vanner , and m. aspelmeyer , nature(london ) * 460 * , 724 ( 2009 ) .
a. xuereb , p. domokos , j. asbth , p. horak , and t. freegarde , phys .
a * 79 * , 053810(2009 ) .
s. mancini , v. giovannetti , d. vitali , and p. tombesi , phys .
. lett . * 88 * , 120401(2002 ) .
w. marshall , c. simon , r. penrose , and d. bouwmeester , phys .
rev . lett.*91 * , 130401 ( 2003 ) .
v. braginsky and s. p. vyatchanin , phys .
a * 293 * , 228 ( 2002 ) .
t. corbitt , y. chen , e. innerhofer , h. muller - ebhardt , d. ottaway , h. rehbein , phys .
lett . * 98 * , 150802 ( 2007 ) .
j. m. courty , a. heidmann , and m. pinard , phys .
90 * , 083601 ( 2003 ) .
d. rugar , r. budakian , h. j. mamin , and b. w. chui , nature ( london)*430 * , 329 ( 2004 ) . i. wilson - rae , n. nooshi , w. zwerger , and t. j. kippenberg , phys . rev . lett.*99 * , 093901 ( 2007 ) .
f. marquardt , j. p. chen , a. a. clerck , and s. m. girvin , phys .
rev . lett.*99 * , 093902 ( 2007 ) .
j. m. dobrindt , i. wilson - rae , and t. j. kippenberg , phys.rev.lett.*101 * , 263602 ( 2008 ) .
z. r. gong , h. ian , yu - xi liu , c. p. sun , and f. nori , phys .
a * 80 * , 665801 ( 2009 ) .
s. huang and g. s. agarwal , phys .
a * 79 * , 013821 ( 2009 ) .
t. kumar , a. bhattacherjee , and manmohan , phys .
a * 81 * , 013835 ( 2010 ) . c. k. law , phys .
a * 21 * , 2537 ( 1995 ) . c. genes , d.vitali , and p. tombesi , new .
j. phys . *
10 * , 095009 ( 2008 ) .
s. mancini , d.vitali , and p.tombesi , phys .
lett . * 80 * , 688 ( 1998 ) .
s. mancini and p.tombesi , phys .
rev.a * 49 * , 4055 ( 1994 ) . v.giovannetti and d.vitali , phys
rev . a * 63 * , 023812 ( 2001 )
. a. imamoglu , h. schmidt , g. woods , and m. deutsch , phys .
79 * , 1467 ( 1997 ) .
p. rabl , phys .
* 107 * , 063601 ( 2011 )
. i. s. gradshteyn and i. m. ryzhik , _ table of integrals , series and products _
( academic press , orlando , 1980 ) ; a. hurwitz , _ selected papers on mathematical trends in control theory _ , edited by r. bellman and r. kabala ( dover , new york , 1964 ) . c. h. metzger and k. karrai , nature(london ) * 432 * , 1002 ( 2004 ) .
j. d. teufel , d.li , m.s .
allman , k. cicak , a. j. sirois , j. d. whittaker , and r. w. simmonds karrai , nature(london ) * 471 * , 204 ( 2011 ) .
m. bhattacharya and p. meystre , phys .
lett . * 99 * , 073601 ( 2007 ) . c. genes , d. vitali , p. tombesi , s. gigan , and m. aspelmeyer , phys.rev.a * 77 * , 033804 ( 2008 ) .
i. wilson - rae , n. nooshi , w. zwerger , and t. j. kippenberg , phys .
. lett . * 99 * , 093901 ( 2007 ) .
f. marquardt , j. p. chen , a. a. clerk , and s. m. girvin , phys .
lett . * 99 * , 093902 ( 2007 ) .
p. f. cohadon , a. heidmann , and m. pinard , phys .
lett . * 83 * , 3174 ( 1999 ) .
e. a. sete and h. eleuch , phys.rev.a * 85 * , 043824 ( 2012 ) .
r. loudon and p. l. knight , j. mod .
* 34 * , 709 ( 1987 ) .
d. vitali , s. gigan , a. ferreira , h. r. bohm , p. tombesi , a. guerreiro , v. vedral , a. zeilinger , and m. aspelmeyer , phys .
* 98 * , 030405 ( 2007 ) .
g. vidal and r. f. werner , phys.rev .
a. * 65 * , 032314 ( 2002 ) ; g. adesso , a. serafini , and f. illuminati , phys . rev .
a. * 70*,022318 ( 2004 ) . | we study theoretically nonlinear effects arising from the presence of a kerr - down conversion nonlinear crystal inside an optomechanical cavity . for this system
we investigate the influences of the two nonlinearities , i.e. , the kerr nonlinearity and the parametric gain , on the dynamics of the oscillating mirror , the intensity and the squeezing spectra of the transmitted field , and the steady - state mirror - field entanglement .
we show that in comparison with a bare optomechanical cavity , the combination of the cavity energy shift due to the kerr nonlinearity and increase in the intracavity photon number due to the gain medium can increase the normal mode splitting in the displacement spectrum of the oscillating mirror and reduce its effective temperature .
our work demonstrates that both the kerr nonlinearity and down conversion process increase the degree of squeezing in the transmitted field .
moreover , we find that in the system under consideration the degree of entanglement between the mechanical and optical modes decreases considerably because of the intracavity photon number reduction in the presence of the kerr medium . |
the conventional approach used to determine decay rates and cross sections is to square the feynman amplitude , average over initial , and sum over final spin states .
this approach is very inconvenient to implement if the number of feynman diagrams and the number of final state particles is large .
this is true even with the use of all the modern tools for symbolic computation .
the conventional approach also has the drawback that in summing over spin states one loses information on spin correlations .
such correlations may be important for exploring new physics at future colliders .
an alternative approach , which has gained popularity in the past decade , is to compute the helicity amplitudes numerically . by doing this ,
spin correlations are easily explored and the process of squaring the amplitude is trivial .
the idea of calculating helicity amplitudes is probably as old as the conventional approach and might even pre - date it@xcite .
more recently , motivated by the need to consider more complicated processes , a number of novel methods@xcite for obtaining helicity amplitudes have been developed .
most of these , however , are viable only in the limit where fermion masses can be ignored . in this limit helicity and chirality are equivalent , leading to enormous simplifications ; this is particularly evident in processes involving the @xmath0-gauge bosons .
most of the recent so - called helicity methods@xcite take advantage of these simplifications@xcite .
although , there have been several attempts to generalize the helicity methods for massless fermions to the massive case , the resulting procedure is rather cumbersome and requires the introduction of arbitrary reference momenta@xcite .
it is the purpose of this paper to present a simple and systematic method for calculating helicity amplitudes which works for processes involving both massless and massive fermions .
the approach that we present leads to a covariant expression for the helicity amplitude in terms of a minimum number of four vector products and is independent of the basis for dirac matrices or specific forms of the spinors .
thus , we believe that our approach leads to an analytic expression which is relatively simple and easy to compare with . for illustration we apply our method to compute the helicity amplitudes for various processes involving the top quark .
we consider the processes : @xmath1
when evaluating a feynman amplitude involving a pair fermions and several gauge bosons , in the initial or final state , the amplitude is expressed as a sum of terms which have the form , @xmath2 where @xmath3 and @xmath4 are the momenta and helicity of the external fermions .
the particular labeling of the indices is chosen for convenience in the manipulations that follow .
note that , as written , ( 1 ) has an even number of @xmath5-matrices sandwiched between spinors .
the case where the number of @xmath5-matrices is odd is treated by setting @xmath6 . in the usual approach
one squares these terms and sums over the fermion spins to obtain a sum over terms of the form , @xmath7.\ ] ] the traces can then be evaluated and the result will be a function of minkowski products of the external particle four momenta .
the expression is then easily integrated over phase space by use of monte carlo methods . in this approach one does not need explicit values of components of free spinors ( which depend on the particular representation used for the @xmath5-matrices ) .
this method is also convenient because , in addition to getting rid of the @xmath5-matrices and expressing the amplitude in terms of simple `` dot '' products , the final result is independent of the unobservable fermion spins . in complicated cases where the amplitude involves the sum over a large number of feynman diagrams , and
each diagram involves many fermion - gauge boson vertices , the number of traces and the length of the traces may render the standard approach impractical .
the standard approach is also inconvenient in cases where interference effects are of interest and one may want to explicitly display the relative phases between the different feynman diagrams . in these cases
it is more convenient to obtain an expression for the amplitude corresponding to each graph which is simple to evaluate numerically .
the square of the amplitude is then obtained by just squaring a single complex number . for processes that only involve massless or nearly massless fermions
there have been many papers where the so called helicity methods for evaluating tree level amplitudes are developed@xcite .
in contrast to the standard approach , the amplitude in the helicity method is expressed in terms of spinor products of the form @xmath8 in this definition the @xmath9 and @xmath10 are light - like four - vectors .
the @xmath11 refer to the helicity of the fermion .
note that @xmath12 , so that up to a phase @xmath13 contains essentially the same information as @xmath14 .
these methods prove relatively simple to implement and lead to compact expressions valid whenever the fermion masses can be ignored .
generalizations of this approach to the massive fermion case@xcite exist , but the results are cumbersome : the method requires the introduction of extra light - like momenta and the expression of the spinor for a massive particle as a sum over spinors for massless particles . consequently , as the number of external particles increases , the number of terms for each diagram grows and results quickly become unmanageable and more prone to human error . in another
approach@xcite some authors found it more convenient to obtain the amplitudes by performing all the @xmath5-matrix products numerically .
although this approach appears to be efficient , the result is not expressible in a simple compact analytic form which one can study .
we finally note that all the approaches quoted above require writing down particular expressions for the spinors .
we will see below that with the approach presented here , it is possible to actually write the amplitude in terms of `` dot '' products without the use of particular expressions for the spinors .
the procedure is relatively simple and leads to compact analytic expressions which are easy to compare with and put into code to evaluate numerically or symbolically .
the basic idea is to observe that ( 1 ) can be re - written as the trace , @xmath15.\ ] ] although this trivial observation has been used by several authors in the past@xcite .
there was no attempt was made to make use of the michael - bouchiat identity ( as we do below ) to obtain an expression for @xmath16 .
the object , @xmath17 , is a four by four matrix in dirac space and can be expressed in terms of an orthogonal basis of the four - dimensional dirac space @xcite .
we do not follow this approach here because we found it simpler to make use of a spinor identity due to michel and bouchiat @xcite , @xmath18 the @xmath19 s are the three pauli matrices and there is an implied sum over @xmath20 .
note that in the case where @xmath21 expression ( [ emb ] ) reduces to the usual projection operator for a state of momentum @xmath9 and helicity @xmath22 .
the @xmath23 s are defined such that , @xmath24 they are the spin vectors corresponding to the four momentum @xmath9 . for , @xmath25 a standard representation of the @xmath23 s is given by , @xmath26 where @xmath27 for the case when @xmath28 .
for completeness we have included a proof of identity ( [ emb ] ) in the appendix .
the identity is only valid if both spinors correspond to the same momenta . to make use of it
we must generalize it to obtain an expression for @xmath16 . if @xmath29 this could be done by applying a boost operator to both sides of ( [ emb ] ) .
it is , however , simpler and more general to observe that any spinor @xmath30 can be expressed in the form @xmath31 where @xmath32 , and @xmath33 is a normalization factor .
this expression is valid whether or not @xmath29 .
note that the helicity index @xmath22 is defined with respect to the direction of , while @xmath34 is defined with respect to ; i.e. the frames in which @xmath22 and @xmath34 are defined are rotated with respect to each other .
multiplying ( 7 ) on the right by @xmath35 and using identity ( 4 ) we obtain , @xmath36 with , @xmath37 it is easy to verify that the matrix @xmath38 is unitary in helicity space as it should be . using ( [ upeqn ] ) in ( 4 )
we obtain the desired generalization , @xmath39 since @xmath40 is unitary it drops out of our expressions upon summing the square of the helicity amplitudes over all helicities . thus ,
if in the end one is going to sum over all helicities one can simplify ( [ meqn ] ) by setting @xmath41 .
the resulting amplitudes will not correspond to the correct helicity amplitudes but the process of squaring them and summing over all helicity indices will indeed generate the correct result for the sum of the squares . if spin correlations are of interest then one needs to make use of the specific expressions for the @xmath40 .
the expressions follow from ( [ clleqn ] ) @xmath42^{1/2 } [ \cos { \frac{\theta}{2 } } \cos { \frac{\theta'}{2 } } + \sin { \frac{\theta}{2 } } \sin { \frac{\theta'}{2 } } { \mbox { $ \mbox{\large e}^{i(\phi ' -\phi)}$ } } ] \nonumber \\[12pt ] c_{-+ } & = & \left [ { \mbox { $ { \displaystyle\frac{ee'+ { \mbox { $ \left|{{\mbox{$\vec{p } $ } } } \right|$ } } { \mbox { $ \left|{{\mbox{$\vec{p}{}\ ' $ } } } \right|$ } } + mm'}{p { \mbox { $ \!\!\cdot\!\!$ } } p'+ mm'}}$ } } \right]^{1/2 } [ \cos{\frac{\theta}{2}}\sin{\frac{\theta'}{2 } } { \mbox { $ \mbox{\large e}^{-i\phi'}$ } } -\sin{\frac{\theta}{2}}\cos{\frac{\theta'}{2 } } { \mbox { $ \mbox{\large e}^{-i\phi } $ } } ] \nonumber \\[12pt ] c_{-- } & = & c_{++}^ * \nonumber \\[12pt ] c_{-+ } & = & -c_{+-}^*.\end{aligned}\ ] ] the angles @xmath43 and @xmath44 correspond to the axial and azimuthal angles for and respectively . in the center of momentum frame where @xmath45 the expressions for @xmath40 simplify to , @xmath46 c_{\pm\mp } &
= & \mp { \mbox { $ \mbox{\large e}^{\pm i \phi}$ } } .\end{aligned}\ ] ] this remains true whether or not @xmath47 .
if in addition it is also true that @xmath47 then ( [ meqn ] ) simplifies to @xmath48 u(p',\lambda ) \otimes \bar{u}(p,-\lambda ) & = & -\lambda\frac{1}{2}\gamma_0({\mbox{$\rlap /\;\!\!p $ } } + m ) ( 1-\lambda\gamma_5 \rlap /\eta^3 ) { \mbox { $ \mbox{\large e}^ { i\lambda\phi}$ } } , \label{meqn2}\end{aligned}\ ] ] where @xmath49 these results also follow by using the relation@xcite @xmath50 in expression ( [ emb ] ) .
finally we remark that in the limit @xmath51 equations ( [ meqn ] ) and ( [ meqn2 ] ) still apply with the proviso that @xmath52 + @xmath53 . in the massless limit , where @xmath54 , expression ( [ meqn ] ) simplifies to @xmath55 { \mbox { $ { \displaystyle\frac{1 } { 2\sqrt{2p { \mbox { $ \!\!\cdot\!\!$ } } p ' } } } $ } } \ , c_{\lambda(-\lambda)}\ , { \mbox{$\rlap /\;\!\!p $ } } ' { \mbox{$\rlap /\;\!\!p $ } } \,\gamma_5\rlap /\eta^\lambda & \mbox{for $ \lambda'=\lambda $ } .
\end{array}\right . \label{meqn3}\ ] ] it is easy to verify that ( [ meqn3 ] ) satisfies all of the identities valid for the inner product of two massless spinors .
for example , @xmath56 and , @xmath57 equation ( [ meqn ] ) , and its simplified forms ( [ meqn2 ] ) and ( [ meqn3 ] ) , play a central role in our method . with them
we can evaluate the trace in ( 3 ) and obtain an analytic expression for the amplitude in terms of dot products . evaluating the trace , even at the amplitude level , may lead to very complicated expressions if the number of external gauge bosons connected to a single fermion line is large .
therefore , we develop an iterative method for obtaining this trace . the iterations can be easily performed either symbolically ( using code like reduce or form ) or numerically .
the basic idea is to use the fact that any product of an odd number of gamma matrices can be written in the form , @xmath58 where the @xmath59 s and @xmath60 s are obtained iteratively from , @xmath61 \nonumber \\ a_n^\mu & = & \frac{1}{4 } tr[\gamma^\mu(\gamma_5\rlap/{v}_{n-1}+{\mbox{$\rlap /{\ !
a}$}}_{n-1 } ) { \mbox{$\rlap /{\ ! b}$}}_{2n}{\mbox{$\rlap /{\
! b}$}}_{2n+1 } ] , \nonumber \\ v_\circ^\mu & = & b_1^\mu \nonumber \\ a_\circ^\mu & = & 0 . \label{ite}\end{aligned}\ ] ] this result follows after repeated use of the @xmath5 identity and @xmath62 . ] , @xmath63 for convenience we define the following expressions , @xmath64 f_{\lambda,\lambda'}(b_0,v_n , a_n)&= & { \mbox{$\bar{u}(p,\lambda ) $ } } { \mbox{$\rlap /{\
! b}$}}_0({\mbox{$\rlap /{\
! v_n}$ } } + \gamma_5 { \mbox{$\rlap /{\ !
a_n}$ } } ) { \mbox{$u(p',\lambda ' ) $ } } .
\label{mfn}\end{aligned}\ ] ] the essence of the method is to write expressions of the form ( [ one ] ) in terms of @xmath65 and @xmath66 defined above .
the @xmath67 s and @xmath68 s are obtained by the simple iteration process ( [ ite ] ) . finally , using ( [ meqn ] ) we express ( [ mfn ] ) in terms of four vector products .
since the general expressions for @xmath65 and @xmath66 are common to every feynman diagram we display their form below .
@xmath69 where @xmath70 & \mbox{for $ \lambda''=\lambda$ } \\[12pt ] - \left [ mm'e^\lambda { \mbox { $ \!\!\cdot\!\!$ } } a + e^\lambda { \mbox { $ \!\!\cdot\!\!$ } } w(p , p',a , v ) \right ] & \mbox{for $ \lambda''=-\lambda,$ } \end{array}\right .
\\[18pt ] n_{\lambda'',\lambda } & = & \left\{\begin{array}{ll } b_0\cdot w(p , p',v , a ) + mm'b_0\cdot v & { } \\[6pt ] \ \ \ \ \ + \lambda\ , b_0 \cdot w(\eta^3,m ' p+ m p',a , v ) & \mbox{for $ \lambda = \lambda$ } \\[12pt ] b_0 \cdot w(\eta^\lambda , m ' p+
m p',a , v ) & \mbox{for $ \lambda '' = -\lambda$ , } \end{array}\right.\end{aligned}\ ] ] and , @xmath71 of course these expressions take on much simpler forms when ( [ meqn2 ] ) and ( [ meqn3 ] ) are applicable .
we will see an example of this simplification in the next section .
we now have at hand all the tools necessary to express any object of the form ( 3 ) in terms of four vector products . in the next section
we illustrate how these techniques are applied to compute the cross section for various processes .
as an illustration of our method we present the helicity amplitudes for two processes involving the top quark , namely , @xmath72 and , @xmath73 these processes will be important in the study of the phenomenology of the top quark at future electron - positron colliders .
we can write the helicity amplitudes for both processes in terms of the spinor objects , @xmath74 we will work in the center of momentum system with the initial particles moving along the @xmath75-axis ; we also take @xmath76 . in this frame
we can then use ( [ meqn2 ] ) to obtain simple expressions for @xmath77 and @xmath78 , @xmath79 b_\mu ( p,\lambda;p',\lambda ' ) & = \begin{cases } -2 m \,g_{o\mu } & \text{if $ \lambda'=\lambda$ } , \\
-2\lambda\ ! { \mbox { $ \left|{\vec{p}}\right|$ } } \!\eta_\mu^\lambda & \text{if $ \lambda'=-\lambda$}. \end{cases } \label{spinfac2 } \end{split}\ ] ] where @xmath80 are the polarization vectors defined in ( [ pol12 ] ) corresponding the momentum @xmath9 , mass @xmath81 and helicity index @xmath22 . [ [ section ] ]
we now have the necessary ingredients to obtain expressions for the helicity amplitudes .
the feynman diagrams for the first process are shown in figure 1 .
we take the @xmath82 , @xmath83 , and @xmath84 couplings to be of the form , and @xmath85 are given by @xmath86 ] , @xmath87 \gamma^\mu_{\gamma e\bar{e } } & = e\gamma^\mu , & \qquad \gamma^\mu_{ze\bar{e } } & = \gamma^\mu(a+b\gamma_5).\end{aligned}\ ] ] the amplitudes can then be written in terms of ( [ spinfac2 ] ) , @xmath88 \\ & \phantom{{=}\frac{1}{s - m_z^2}\biggl\ { } + ad\ , \bigl[a_\mu ( \lambda_e,\lambda_{\bar{e}}){}^\dagger\ , b^\mu ( \lambda_t,\lambda_{\bar{t}})\bigr ] \\ & \phantom{{=}\frac{1}{s - m_z^2}\biggl\ { } + bc\,\bigl[b_\mu ( \lambda_e,\lambda_{\bar{e}}){}^\dagger\ , a^\mu ( \lambda_t,\lambda_{\bar{t}})\bigr ] \\ & \phantom{{=}\frac{1}{s - m_z^2}\biggl\ { } + bd\,\bigl[b_\mu ( \lambda_e,\lambda_{\bar{e}}){}^\dagger\ , b^\mu ( \lambda_t,\lambda_{\bar{t}})\bigr ] \biggr\}. \end{split } \label{tamp}\ ] ] for simplicity we have suppressed the momentum indices , but the momentum dependence should be clear from the context .
the electron and positron momenta are designated by @xmath89 and @xmath90 , and the top , anti - top quark momenta are designated by @xmath9 and @xmath91 . in the cm frame
the momenta take the form , @xmath92 where @xmath93 is the energy of the incoming electron in the center of mass frame .
the electron mass is ignored in all of our expressions . using these expressions for the momenta together with
( [ spinfac2 ] ) and ( [ tamp ] ) it is simple to obtain an expression for the helicity amplitudes .
we represent the total amplitude by , @xmath94 , where the helicity index for each particle is manifest . then , @xmath95 \\[6pt ] t(\lambda_e,-\lambda_{e};\lambda_t,\lambda_{t } ) & = -\frac{m_t\,\lambda_t}{e\,\beta_z^2}\ , \bigl [ \,e\,q_t\,\beta_z^2 + ( a+b\lambda_e)c\ , \bigr ] , \end{split}\ ] ] where @xmath96 , and @xmath97 . in obtaining
the above expression the following relations were useful , @xmath98 [ [ section-1 ] ] as another example we consider the process @xmath99 which will be present in the back - scattered laser mode of an @xmath100 collider .
this experiment will have the possibility of polarizing both photons and of analyzing the polarization of the final state quarks .
the relevant diagrams are those obtained via a @xmath101-quark exchanged in the @xmath101-channel ( fig .
[ fig2 ] ) . a straightforward application of ( [ meqn ] ) or ( [ mfn ] ) in the cm frame ( [ cmvecs ] ) leads to the following expression for the amplitude @xmath102 , @xmath103 \vartheta_1^i \vartheta_2^j\ ] ] where @xmath104 denote the top and anti - top helicities and @xmath105 denote the photon polarization vectors , @xmath106 denoting by @xmath107 the photon helicities .
we also defined @xmath108 \\
t_{(\pm)}^{ij } ( h_1,h_2,\lamt,-\lamt ) & = 2e\,\biggl [ p^i \left ( \eta^\lamt \right)^j + p^j \left ( \eta^\lamt \right)^i \pm\vec{k}_1\cdot\vec{\eta}^\lamt\,\delta^{ij}\biggl ] \end{split}\ ] ] in terms of the momenta in ( [ cmvecs ] ) we obtain , @xmath109 t ( h_1,-h_1,\lamt , \lamtbar ) & = q_t^2\ , \frac{2\,\beta_t\,\lamt\sin\theta}{1-\beta_t^2\cos^2{\theta } } \begin{cases } { \mbox { $ { \displaystyle\frac{m_t}{e}}$ } } \sin{\theta } & \text{if $ \lamtbar=\lamt$},\\[10pt ] ( h_1+\lamt\cos{\theta } ) & \text{if $ \lamtbar=-\lamt$}. \end{cases } \end{split}\ ] ] here @xmath110 denotes the velocity of the top - quark , @xmath111 , and @xmath112 is the cm scattering angle . note that as @xmath113 , @xmath114 as required by helicity conservation .
also in the non - relativistic limit the amplitude for two photons into two like - spin electrons vanishes , as it should .
we have presented a new method for computing helicity amplitudes which is applicable for processes involving both massless and massive fermions .
the method allows one to obtain a covariant expression for the helicity amplitude without having to go through the process of squaring the amplitude .
the method can be easily adopted to both numerical and symbolic computations .
we presented two examples which illustrate the simplicity of the method .
although for our two examples we worked in the center of mass frame the method is applicable in any frame and can be used in processes with many pairs of external massive fermions . in particular
the methods presented here were used to compute the cross sections for @xmath115 , @xmath116 , @xmath117 , and many others ( see for example references @xcite ) .
for very complicated cases the combination of this method with a program for algebraic manipulation ( such as form , reduce , macyma , maple , or mathematica ) should greatly simplify calculations . even in those cases where one is only interested in the total cross section
the use of the methods presented here may , in some cases , prove simpler to use than the traditional approach .
r.v . would like to thank howard haber and michael peskin for very helpful discussions .
this research was supported in part by doe contracts de - fg05 - 92er40722 and de - fg03 - 92er40837 and by the lightner - sams foundation of dallas .
in this appendix we present a proof of the michel - bouchiat identity ( [ emb ] ) .
we follow here the notation of reference .
the state of a particle can be characterized by giving the momentum @xmath118 , mass @xmath81 , total spin @xmath119 , and its component of spin along a fixed direction .
it is common to take the @xmath75-axis or the direction of the momentum to specify this fixed direction .
the latter case is refered to as the helicity basis . in what follows
we will use the helicity basis .
the advantages of using the helicity basis are well known@xcite and we will not dwell on them here .
the helicity operator is defined by , @xmath120 where @xmath121 is the pauli - lubanski four vector ; @xmath122 and @xmath123 are the angular momentum and energy - momentum operators respectively .
the commutation relation for the @xmath0 s is , @xmath124=-i\epsilon^{\mu\nu\rho\sigma}\,p_\rho w_\sigma.\ ] ] this relation follows from the commutation properties of @xmath122 and @xmath123 .
the four vector @xmath125 is any space - like normalized four vector with @xmath126 . in the helicity basis @xmath127
is in the direction of @xmath118 .
a conventional form for @xmath125 is , @xmath128 following michel@xcite we introduce an orthogonal set of tensors , @xmath129 such that , @xmath130 if we define , @xmath131 , then it follows that , @xmath132 . for a particle of spin @xmath119 , with @xmath133 , the operator , @xmath134 ,
has @xmath135 eigenvalues : @xmath136 .
these eigenvalues are used to label states of given momenta and spin . from the commutation relation for the @xmath0 s and the properties of the @xmath23 s it is straightforward to show that the commutation relation for the @xmath137 s obey an @xmath138 algebra , @xmath139= i\epsilon_{ijk}s_k,\ ] ] furthermore
, if we pick @xmath140 to have the form , @xmath141 then up to a factor of @xmath142 , the operator @xmath143 , is just the helicity operator . in the helicity basis states
are labeled by the eigenvalues of @xmath143 .
the helicity projection operator is given by , @xmath144 = & \frac{1}{2 } \left(1\pm 2s_3\right),\end{aligned}\ ] ] and as usual the lowering and raising operators of helicity are given by , @xmath145 to obtain the michel - bouchiat form ( [ emb ] ) we recall that , @xmath153 and use the definition of @xmath154 to obtain , @xmath155 and , @xmath156 u(p,\mp)\,\bar{u}(p,\pm ) & = & \gamma_5\frac{\rlap/ \eta_1 \mp i\rlap/ \eta_2}{2 } ( { \mbox{$\rlap /\;\!\!p $ } } + m).\end{aligned}\ ] ] finally using the pauli matrices , @xmath157 , these expressions can be written in the shorthand notation of ( [ emb ] ) . , phys .
105b * , 215 ( 1981 ) ; phys . lett . *
114b * , 203 ( 1982 ) ; nucl . phys . *
b206 * , 53 ( 1982 ) ; nucl . phys . *
b206 * , 61 ( 1982 ) ; nucl .
b239 * , 382 ( 1984 ) ; nucl . phys . *
b239 * , 395 ( 1984 ) .
j. f. gunion and z. kunszt , phys .
b161 * , 333 ( 1951 ) . | we present an alternative approach for calculating helicity amplitudes for processes involving both massless and massive fermions . with this method one can easily obtain covariant expressions for the helicity amplitudes .
the final expressions involve only four vector products and are independent of the basis for gamma matrices or specific form of the spinors .
we use the method to obtain the helicity amplitudes for several processes involving top quark production .
smu - hep-94/28 + ucrhep - t152 + * a covariant method for calculating helicity amplitudes * roberto vega + _ department of physics _ + _ southern methodist university _
+ _ dallas , tx 75275 _ jose wudka + _ department of physics _
+ _ university of california at riverside _ + _ riverside , ca 92521 - 0413 _ |
seventy years after pauli s @xcite proposal of its existence and almost half a century after its discovery @xcite , the neutrino still plays a central role in elementary particle physics .
the main problem of determining whether it is a massive or massless particle seems to have been solved after the last evidences , coming mainly from the solar and atmospheric neutrino experiments , but we still have to answer important questions .
we do nt know , for instance , whether it is a majorana or dirac particle , nor have we a unique natural explanation of its lightness @xcite . the problem of searching for neutrino mass and studying the oscillation phenomenon has been faced in the past through many different experimental techniques .
the first studies , based on the so called fermi - perrin method of the observation of the @xmath5 spectrum near the end point , gave @xcite the limit @xmath6 .
this limit was obviously lowered many times in the following years , up to the present results @xcite . since the experiment of goldhaber et al .
( 58 ) @xcite , we know that neutrinos produced in @xmath5 decays are left - handed particles .
this fact at the beginning appeared as a confirmation of the hypothesis that the neutrino is a massless particle .
another milestone in the development of our knowledge of neutrino physics was the idea suggested by pontecorvo @xcite that neutrino can `` oscillate '' , in the sense that the flavor states are superposition of different mass states .
this was a revolutionary hypothesis , because only the electronic neutrino was known in those days , but nowadays we have strong experimental hints that would confirm the validity of pontecorvo s idea .
coming to our days , in the usual version of the standard model that describes the electroweak interactions , the neutrino is a left handed dirac particle ; hence , in such a theory it is impossible to build a renormalizable mass term for this particle . on the other hand , as we are going to see in detail , there is experimental evidence that it is a massive and oscillating particle .
therefore we are forced by the data to enlarge this `` minimal version '' of the standard model and possibly to build a more general theory in which a neutrino mass can be fitted naturally .
neutrino physics can be considered an ideal playground to test different theories beyond the standard model , like , for instance , supersymmetry and grand unification theories @xcite or the theories based on the existence of large extradimensions @xcite .
the determination of the value of neutrino mass also has important implications on cosmological models .
in particular neutrino is a candidate for dark matter and this fact determined a revival of astrophysical studies of neutrino properties in the last decade @xcite .
all the experiments aiming to measure the neutrino mass and to test the existence of oscillations can be classified in some main categories .
first of all there are the direct kinematical searches like the ones of @xcite and the searches for the neutrinoless double @xmath5 decays @xcite .
the present limits on the values of @xmath7 and @xmath8 masses are @xcite : @xmath9 the best limits for the mass of the electron neutrino , instead , have been obtained from the mainz and the troitsk @xcite experiments which have found @xmath10 . in future many experiments will try to lower this limit .
in particular there is a great expectation for katrin ( the karlsruhe tritium neutrino experiment ) @xcite , that should start data taking in 2007 and improve the sensitivity down to @xmath11 .
the search for neutrinoless double @xmath5 decays is important because the observation of these decays would be a clear indication in favor of a majorana nature of the neutrino , if we assume cpt invariance @xcite .
the most stringent limit on this process available at the moment comes from the heidelberg - moskow collaboration @xcite @xmath12 and from igex ( international germanium experiment ) @xcite @xmath13 . in the last year
there has been a claim @xcite from some members of the heidelberg - moskow collaboration of discovery of a @xmath14 effect that would be a signal of neutrinoless double @xmath5 decay , but this result has been strongly contested and the discussion on its validity is still an open question .
a second group of experiments uses neutrino fluxes produced at accelerators and nuclear reactors .
they are usually divided in long- and short - baseline , according to the distance between the neutrino production point and the detector .
many short baseline accelerator experiments did nt find any signal of oscillation .
they are nevertheless important , because they give constraints on the possible values of the mixing parameters .
the most important limits have been obtained by nomad @xcite and chorus @xcite at cern .
these two experiments were designed to check relatively high values of the mass differences ( @xmath15 ) and used a beam of @xmath8 to look for a signal of a @xmath16 production , that would have been an indication of @xmath17 oscillations . besides the long baseline reactor experiments it is worthwhile to recall the results of chooz @xcite and palo verde @xcite . at chooz
a beam of reactor @xmath18 was sent to a detector located about 1 km away and detected through the reaction @xmath19 .
no evidence of oscillation was found at chooz and the experimental result for @xmath20 , that is the ratio between the number of measured @xmath18 events and the expected number in absence of oscillation , is compatible with @xmath21 . in a simple two flavor model
the oscillation probability is given by the relation @xmath22 where @xmath23 is the distance source - detector expressed in meters , e is the @xmath24 energy in and @xmath25 is the difference of the squares of neutrino masses expressed in @xmath26 .
the range of mass differences and mixing angles that can be tested in a certain experiment is limited by the requirements that the source - detector distance be much shorter than the oscillation length @xmath27 the chooz average energy value is @xmath28 ; therefore chooz results can be used to exclude a significant part of the mixing parameters plane . in particular
they tell us that @xmath25 must be smaller than @xmath29 , unless the values of the mixing angle are very small .
the opposite situation took place in the case of lsnd @xcite , a short baseline accelerator experiment performed with a neutrino beam produced at the los alamos meson physics facility ( lampf ) .
the experiment found evidence of two kinds of oscillation signals .
the first was the excess of @xmath18 in the beam of @xmath30 produced by the decay at rest of the @xmath31 obtained as secondary products of the proton accelerator beam .
the second was a signal of @xmath32 oscillations , starting from the @xmath8 produced by the @xmath33 decay in flight .
the lsnd result , if confirmed , ought to be a clear indication of oscillation with very high values of the mass difference , up to @xmath34 . to reconciliate this result with the ones coming from solar and atmospheric neutrinos
one would have to postulate the existence of at least one sterile neutrino in addition to the usual three active ones .
however , up to now there has been no independent confirmation of the lsnd results .
the karmen experiment @xcite , performed at the rutherford laboratories , explored a significant part of the mixing parameter space proposed by lsnd and it did nt find any signal of oscillation . a new experiment miniboone @xcite is going to run very soon and produce data starting from 2004 .
it is very similar to lsnd and will test definitely the validity of lsnd results . a new generation of very long baseline experiments has become available in the last years . the forerunner of them is k2k
@xcite , that uses a neutrino beam produced at the japan kaon facility kek and detected at the kamioka site . up to
now k2k has detected 56 events instead of the expected value in absence of oscillations of @xmath35 events .
this is a confirmation of neutrino oscillations ( the no - oscillation probability is less than @xmath36 ) .
moreover the best fit point @xcite values for the mass difference and the mixing angle ( @xmath37 and @xmath38 ) are in good agreement with the results of atmospheric neutrino experiments .
two similar projects have already been approved and will become available in the near future : one of them is a neutrino beam from cern to the gran sasso labs @xcite and the other one is in the usa @xcite ( from fnal to soudan ) .
the long baseline accelerator experiments will probably give an important confirmation of the oscillation evidence which have up to now come from the study of solar and atmospheric neutrinos .
they are also expected to find in an unambiguous way indications of oscillation from appearance signals .
in addition , in the long baseline experiment one has the opportunity of choosing the specific characteristic of the beam ; hence they can be used to perform precision measurements @xcite .
for instance they should be useful to study the value of the mixing angle @xmath39 , relevant for eventual cp violation .
the present limit on the measurement of this angle coming from chooz ( @xmath40 degrees ) , could be lowered to the level of about @xmath41 degrees at icarus , one of the two experiments that will use the cern - gran sasso beam .
important results should very soon come from the long baseline reactor experiment kamland @xcite , which might in principle give a definite solution to the solar neutrino problem , as we will see in the following .
the two main categories of experiments looking for oscillation signals are the ones that study the atmospheric and the solar neutrinos .
the atmospheric neutrinos are products of decay of the cosmic rays .
the number of electronic and muonic neutrinos can be computed with good accuracy , considering the properties of cosmic rays , their decay channels and eventually geomagnetic effects .
most of the atmospheric neutrino experiments measure the value of the double ratio @xmath42 the numerator and denominator are respectively the experimental and the monte carlo computed values of the ratio between the events generated by muonic neutrinos ( and antineutrinos ) and the ones generated by electronic neutrinos ( antineutrinos ) .
there are essentially two kind of experiments : the water cherenkov ( like kamiokande @xcite , super - kamiokande @xcite , imb @xcite ) and the iron plate calorimeters ( like soudan ii @xcite and in the past years frejus @xcite and nusex @xcite ) .
clear evidence of oscillations has been found at kamiokande , super - kamiokande ( sk ) , imb and soudan ii and also at the macro @xcite experiment at gran sasso . the best statistic has been obtained at sk , which found @xcite @xmath43 for the sub - gev events and @xmath44 for the multi - gev events .
another interesting observable is the up - down asymmetry between the up going events , in which the neutrino crossed the earth before interacting in the detector , and the down going ones : @xmath45 .
the experimental value of this quantity is consistent with zero for the electronic neutrinos , while for the muonic ones the up - down asymmetry for high values of the momenta is a decreasing negative value .
these results are clear indications of a reduction of the flux of muonic neutrinos and antineurinos that arrive at the detector after crossing the earth .
the most natural explanation of this phenomenon is the possibility that the muonic neutrinos oscillate into other flavors and the oscillation probability is greatly enhanced by the interaction with matter .
the last group of experiments is that of the experiments observing the neutrinos coming from the sun . we will discuss them in detail in the rest of the paper . the first experiment on solar neutrinos , homestake @xcite , started at the end of the @xmath46 using the inverse @xmath5 decay on chlorine @xmath47
the threshold energy was @xmath48 , hence it was sensitive to the pep , @xmath49 , @xmath50 and hep components of the solar neutrino flux .
the results were really surprising , because homestake found a deficit of the solar neutrino flux of more than @xmath51 that predicted by the solar standard model ( ssm ) .
the updated value of the ratio @xmath20 , between the experimental results and the ssm prediction @xcite , for the chlorine experiment is @xmath52 .
this result raised fundamental questions : what happens to solar @xmath24 on their way to earth ?
eventually , could the ssm be wrong ?
the homestake indication was confirmed by similar experiments , sage @xcite in russia and gallex @xcite and later on gno @xcite at the infn gran sasso labs , which used gallium instead of chlorine . the energy threshold is lower in the gallium experiments ( @xmath53 ) making them also sensitive to pp neutrinos , which are the main component of the solar neutrino flux .
the updated gallium results are @xcite : @xmath54 this confirmation of homestake results gave a strong support to the neutrino oscillation hypothesis and caused an increase of the interest for this problem .
it could be a signal of _ new physics_. an essential improvement in the knowledge of solar neutrinos came with the advent of the water cherenkov experiments , kamiokande @xcite and super - kamiokande ( sk ) @xcite , that looked at the elastic scattering @xmath55 and confirmed the existence of the `` solar neutrino problem '' with a very high statistic . in this
experiments it was possible to know the direction of the incoming neutrino ( by looking at the outgoing direction of the recoil electron ) and also to study the energy and angular spectrum and the day night asymmetries .
the energy threshold for these experiments was quite high ( 5 for sk ) and therefore they were sensitive only to the high energy component of the neutrino flux , that is @xmath50 and hep neutrinos .
their results confirmed the existence of a deficit in the electron neutrinos reaching the detector .
the sk result for the energy spectrum and the small values of the day - night asymmetries were also very important to put strong constraints on the possible values of the mixing parameters .
after the publication of sk data it was clear that there was a deficit of solar electron neutrinos reaching the earth , with respect to the flux predicted by ssm .
the oscillation hypothesis was considered the most plausible explanation of this phenomenon , but there were still different regions allowed by the experiments in the mixing parameter plane , as we will see in detail .
the real breakthrough was due to the sno experiment that published its first data in 2001 @xcite .
sno is a deuterium cherenkov detector designed to look simultaneously at three different reactions : @xmath56 the first reaction ( cc ) receives contribution only from the electron neutrino , while the others ( nc and es ) are sensitive to all neutrino flavors .
this experiment gives the first direct model independent measurement of the total solar neutrino flux reaching the earth ( through the nc observation ) and at the same time , comparing this flux with the one of @xmath57 recovered from cc , it offers a strong evidence of the oscillation of @xmath57 into other active neutrinos . during its first phase of working @xcite sno
observed the charged current and elastic scattering events , with an energy threshold for electron detection of 6.75 . the @xmath57 flux measured from cc , after 241 days of running , was : @xmath58 .
the ratio between this value and the ssm prediction was @xmath59 . in the sno experiment
the neutrino flux can be recovered also from the elastic scattering , using the relation : and @xmath7 to the elastic scattering cross section is only through neutral current , hence it is about 1/6 of the contribution of @xmath57 that can interact also through charged current ] @xmath60 the value of the total neutrino flux recovered from the elastic scattering at sno and also , with a better statistics , at sk does nt agree with the @xmath57 flux obtained from sno cc .
the comparison of the two results gives : @xmath61 this result was the first evidence ( at @xmath62 level ) of the presence of muonic and tauonic neutrinos in a electronic neutrino beam reaching the earth from the sun .
therefore it was , up to the present sno data on nc , the most robust evidence of @xmath57 oscillation into other active neutrinos .
it is also remarkable that the sum of the @xmath57 and @xmath63 fluxes give a value in good agreement with the ssm prediction .
consequently the results of sno phase i also strongly disfavored the hypothesis of pure oscillation into sterile neutrinos .
recently , while this paper was in preparation , the data of the so called phase ii of sno also became available @xcite .
this data , obtained with 306.4 days of running , confirms the indications of the phase i and includes the first neutral currents ( nc ) observations .
given all the experimental data that we have just reported , one can say that there is really strong evidence that neutrinos are massive and oscillating particles .
nevertheless , many details of the mass patterns still have to be clarified . with this aim in mind
, we developed a global analysis of all the available data on solar neutrinos , also including the chooz constraints .
our first purpose was that of determining the regions in the mixing parameter plane that are still compatible with the experiments .
in addition to this , we wanted to understand how the forthcoming experiments ( in particular borexino and kamland ) could improve our knowledge of neutrino mass properties .
we assumed neutrino oscillation as a working hypothesis and considered bidimensional models .
for details of our analysis we refer the interested reader to @xcite . here
we just report the most salient aspects of our strategy .
the analysis is based on the numerical calculation of the expected event rate for every solar neutrino experiment as a function of the mixing parameters and on the comparison between these expected numbers and the experimental data .
the statistical analysis is based on the @xmath64 method its output being contour plots in which one can see which values of the mixing angles and mass differences are still allowed at a given confidence level .
our calculation can essentially be split into two parts .
the first one is the determination of the neutrino transition amplitude , i.e. the probability for an electronic neutrino produced in the sun to change its flavor before reaching the detector .
the other ingredient is the calculation of the detector response functions , that , for a given neutrino energy , depend on the experimental details of the specific detector ( i.e. efficiency , resolution , etc . ) and on the cross section for the reaction under examination .
the transition amplitude calculation is separated in three parts , corresponding to the neutrino propagation inside the sun , in the vacuum and in the earth .
for every value of the mixing parameters we compute fully numerically the amplitudes in the sun and in the earth , while the one corresponding to the vacuum evolution is computed analytically .
the three amplitudes are patched together using the evolution operator formalism @xcite .
we included in our analysis the total rates of the chlorine and gallium experiments , together with the different energy bins of superkamiokande and with the charged current results of the first phase of sno .
the resulting contour plots are reported in figure 1 .
one can note that there are still two allowed region , even at @xmath65 c.l . : the large mixing angle ( lma ) , where we found the best fit point with a goodness of fit ( g.o.f . ) of @xmath66 , and the so called low solution , characterized by lower values of the mass differences .
the two other possible solutions of the solar neutrino problem , that historically have been considered , are the small mixing angle ( sma ) and the vacuum solution , corresponding to much lower values of @xmath25 .
these two solutions can not be completely excluded , even if they are strongly disfavored mainly by sk data on the day and night energy spectrum and by the sno results .
another effect of sno data was that of shrinking the different regions which became well separated .
our results are in good agreement with most of the analysis one could find in literature @xcite . and @xmath67 . _
given this situation , we studied which new information should come in future from the borexino data @xcite .
borexino @xcite is a solar neutrino experiment , mainly sensitive to the @xmath49 component of the neutrino flux , that should start running in very next years at the gran sasso labs . in figure 2 the usual contour plots obtained from all the experiments available up to now
are superimposed to the contour lines corresponding to different hypothetical possible values of the total rate at borexino .
as one can see from the picture , borexino should be able to clarify the situation in the case in which the solution very well is in the small mixing angle region .
the situation would be , instead , more complicate in case of lma or low solutions .
in these two regions , in fact , the ratio between the borexino signal and the ssm prediction in absence of oscillations should be between 0.6 and 0.7 .
the discrimination power of borexino increases a lot if we look also at the day - night asymmetry , as one can see from figure 3 .
the low region is characterized by high values of the asymmetry , that can reach up to @xmath68 , while in the lma region the day - night asymmetry is much lower .
hence , by looking simultaneously at the total rate and at the day - night asymmetry , borexino should be able to discriminate between the two solutions of the solar neutrino problem that are compatible with the experiments up to now , that is the lma and the low solutions .
another very important experiment , already running , that should significantly improve our knowledge of the mixing parameters relevant for solar neutrinos is kamland @xcite . in this experiment a flux of low energy @xmath18 produced by different nuclear reactors
is sent to a scintillator detector capable of detecting their interactions with protons .
although it is not a traditional solar neutrino experiment , kamland is sensitive to neutrino oscillations with mixing parameters in the lma region , that seems to be the solution of the solar neutrino problem preferred by the present data .
therefore , we can hope that kamland will soon be able to determine the exact values of the mixing parameters with satisfactory accuracy .
the main limitation of kamland is its reduced sensitivity to the extreme upper part of the lma region , that could create problems in the determination of @xmath69 , as discussed in @xcite and later on in @xcite . for a detailed discussion about kamland potentiality and discrimination power
we refer the interested reader to @xcite .
while this paper was in preparation the data of the second phase of sno was published @xcite .
they contained the first direct oservation of the neutral current ( nc ) process and the data of the cc and es processes with statistics higher than the one the first sno phase @xcite . from the nc data
one can recover a value of the total active flux @xmath70 which is in very good agreement with the ssm prediction ( @xmath71 , if we use the old value for @xmath72 ) .
this result is an important confirmation of the validity of the ssm . at the same time , comparing these values with the sk and sno es values of the @xmath50 electron neutrino flux one obtains proof that a significant part of the electron neutrinos coming from the sun is converted into other active flavors ( there is a @xmath73 evidence that the flux of @xmath63 is different from zero ) .
we have redone our analysis with the addition of these recent sno data . in @xcite
we have assumed the simplifying hypothesis that the spectrum is undistorted with respect to the form predicted by the ssm in absence of oscillations .
this assumption is essentially valid in the lma region ( the one preferred by the data at the moment ) . for the detailed values of the mixing parameters we recovered in this region and for the related study of kamland potentiality we refer the reader to @xcite .
here we just recall that our results are in good agreement with other similar analysis @xcite .
we are also doing a more sophisticated analysis of the full mixing parameter plane , without the undistorted spectrum hypothesis and any other model dependent assumption . a critical analysis of the influence of the different experimental results and of the possible experiments that should come after borexino and kamland is performed , for instance , in @xcite
one of us v.a . would like to thank the organizers of the conference `` les rencontres de physique de la vallee daoste '' for the invitation and the organizers and participants of the conference for the nice and stimulating atmosphere .
we all are grateful to prof .
s. petcov for useful and interesting discussions .
99 see : w. pauli , in neutrino physics ( 1991 ) , pag . 1 , edited by k. winter , cambridge university press c. l. cowan , f. reines , f. b. harrison , h. w. cruse and a.d .
mcguire , science * 124 * , 103 ( 1956 ) for a review on neutrino mass and oscillations see , for instance : s. m. bilenky , c. giunti and w. grimus , progress in part . and nucl .
* 43 * , 1 ( 1999 ) ; s. m. bilenky , arxiv : hep - ph/0205047 , talk given at the xvi rencontres de physique de la vallee daoste , la thuile , italy , march 2002 ; g. altarelli and f. feruglio , arxiv : hep - ph/0206077 , to appear in `` neutrino mass '' , springer tracts in modern physics , ( ed . by g. altarelli and k. winter ) ;
e. k. akhmedov , nucl .
* 100 * , 295 ( 2001 ) .
e. k. akhmedov , arxiv : hep - ph/0001264 .
o. kofoed - hansen , phys . rev . * 71 * , 451 ( 1947 ) ; g. c. hanna and b. pontecorvo , phys . rev . * 75 * , 451 ( 1949 ) r. barate
_ et al . _ ,
j. * c2 * , 395 ( 1998 ) ; see also : + j. m. roney , nucl .
. suppl . *
91 * , 287 ( 2001 ) d. e. groom _ et al .
j. * c15 * , 1 ( 2000 ) j. bonn _
et al _ , proc.of the int .
neutrino 2000 , sudbury , canada , nucl . phys .
* b * proc .
91 , 273 ( 2001 ) m. goldhaber , l. grodzins and a.w .
sunyar , phys . rev . * 109 * , 1015 ( 1958 ) b. pontecorvo , j. exptl .
* 33 * , 549 ( 1957 ) [ sov .
jetp * 6 * , 429 ( 1958 ) ] ; b. pontecorvo , j. exptl . theoret .
phys . , * 34 * , 247 ( 1958 ) [ sov .
jetp * 7 * , 172 ( 1958 ) ] see for instance : t. fukuyama , n. okada , arxiv : hep - ph/0205066 ; r. kitano , arxiv : hep - ph/0204164 ; b. bajc , g. senjanovic and f. vissani , arxiv : hep - ph/0110310 see for instance : n. arkani - hamed , s. dimopoulos , g. r. dvali and j. march - russell , phys .
* d * 65 , 024032 ( 2002 ) ; r. n. mohapatra , s. nandi and a. perez - lorenzana , phys . lett .
* b 466 * , 115 ( 99 ) ; r.barbieri , p. creminelli and a. strumia , nucl . phys .
* b 585 * , 28 ( 2000 ) ; a. lukas , p. ramond , a. romanino and g. g. ross , phys . lett . * b 495 * , 136 ( 2000 ) ; h. davoudiasl , p. langacker and m. perelstein , phys .
rev . * d 65 * , 105015 ( 2002 ) ; k. r. dienes , e. dudas and t. gherghetta , nucl .
* b 557 * , 25 ( 99 ) see for instance : a. d. dolgov , neutrinos in cosmology , arxiv : hep - ph/0202122 and r. n. mohapatra and p.b .
pal , massive neutrinos in physics and astrophysics , world scientific lecture notes in physics , vol .
41 , ( world scientific , singapore , 1991 ) h. v. klapdor - kleingrothaus _ et al .
_ , eur .
j. a * 12 * , 147 ( 2001 ) . c. e. aalseth _
[ 16ex collaboration ] , phys .
d * 65 * , 092007 ( 2002 ) [ arxiv : hep - ex/0202026 ] .
j. bonn _ et al .
_ , nucl .
suppl . * 91 * , 273 ( 2001 ) ; ch .
et al . _ , phys .
b460 * , 219 ( 1999 ) v. m. lobashev _ et al .
_ , nucl .
suppl . * 91 * , 280 ( 2001 ) ; v. m. lobashev _ et al .
_ , phys . lett .
* b460 * ( 1999 ) a. osipowicz _
et al _ , katrin letter of intent , arxiv : hep - ex/0109033 h. v. klapdor - kleingrothaus , a. dietz , h. l. harney and i. v. krivosheina , mod .
a * 16 * , 2409 ( 2001 ) [ arxiv : hep - ph/0201231 ] .
s. pascoli and s. t. petcov , arxiv : hep - ph/0111203 .
j. altegoer et al . , nomad coll . , phys .
* b 431 * , 219 ( 1998 ) chorus coll .
, arxiv hep - ex/9807024 m. apollonio et al . , chooz coll . , phys .
* b 466 * , 415 ( 1999 ) y. f. wang [ palo verde collaboration ] , int .
j. mod .
a * 16s1b * , 739 ( 2001 ) .
f. boehm _ et al .
_ , phys .
d * 64 * , 112001 ( 2001 ) [ arxiv : hep - ex/0107009 ] .
m. sung , int .
j. mod .
a * 16s1b * , 752 ( 2001 ) . c. athanassopoulos _ et al .
_ , lsnd coll .
81 * , 1774 ( 1998 ) a. strumia , arxiv : hep - ph/0201134 . j. wolf [ karmen collaboration ] , k. eitel and b. zeitniz , karmen coll .
, arxiv : hep - ex/9809007 g. drexlin , talk given at _
neutrino 2002 _ , xxth international conference on neutrino physics and astrophysics , may 2002 , munich e. a. hawker , int .
j. mod .
a * 16s1b * , 755 ( 2001 ) .
r. stefanski [ miniboone collaboration ] , nucl .
suppl . * 110 * , 420 ( 2002 ) .
r. j. wilkes [ k2k collaboration ] , `` results from k2k '' , published in venice 2001 , neutrino telescopes , vol . 1 179 - 190 . c. yanagisawa , nucl .
suppl . *
95 * , 130 ( 2001 ) .
k. nishikawa , for the k2k collaboration , talk given at _
neutrino 2002 _ , xxth international conference on neutrino physics and astrophysics , may 2002 , munich s. katsanevas , the cngs program status and physics potential , talk given at _
neutrino 2002 _ , xxth international conference on neutrino physics and astrophysics , may 2002 , munich a. weber [ minos and opera collaborations ] , j. rico [ icarus collaboration ] , arxiv : hep - ex/0205028 .
f. arneodo _ et al .
_ , nucl .
instrum .
a * 461 * , 324 ( 2001 ) .
v. paolone nucl .
suppl . * 100 * , 197 ( 2001 ) .
k. lang [ minos collaboration ] , nucl .
instrum .
a * 461 * , 290 ( 2001 ) .
t. ota and j. sato , arxiv : hep - ph/0202145 .
v. d. barger , a. m. gago , d. marfatia , w. j. teves , b. p. wood and r. zukanovich funchal , phys .
d * 65 * , 053016 ( 2002 ) [ arxiv : hep - ph/0110393 ] .
a. piepke [ kamland collaboration ] , nucl .
. suppl . *
91 * , 99 ( 2001 ) ; p. alivisatos _
et al _ , stanford - hep-98 - 03 j. shirai , start of kamland , talk given at _
neutrino 2002 _ , xxth international conference on neutrino physics and astrophysics , may 2002 , munich s. hatakeyama _ et al . _
[ kamiokande collaboration ] , phys .
lett . * 81 * , 2016 ( 1998 ) [ arxiv : hep - ex/9806038 ] . y. oyama , phys .
d * 57 * , 6594 ( 1998 ) .
y. fukuda _ et al . _
[ kamiokande collaboration ] , phys .
b * 335 * , 237 ( 1994 ) .
k. s. hirata _ et al . _
[ kamiokande - ii collaboration ] , phys .
b * 280 * , 146 ( 1992 ) .
y. fukuda _ et al .
_ [ super - kamiokande collaboration ] , phys .
* 81 * , 1562 ( 1998 ) [ arxiv : hep - ex/9807003 ] .
t. kajita [ super - kamiokande collaboration ] , nucl .
suppl . * 100 * , 107 ( 2001 ) .
t. toshito [ superkamiokande collaboration ] , arxiv : hep - ex/0105023 . s. fukuda _ et al .
_ [ super - kamiokande collaboration ] , phys . rev .
* 85 * , 3999 ( 2000 ) [ arxiv : hep - ex/0009001 ]
. m. shiozawa , for the super - kamiokande collaboration , talk given at _
neutrino 2002 _ , xxth international conference on neutrino physics and astrophysics , may 2002 , munich r. becker - szendy _ et al .
_ , nucl .
suppl . * 38 * , 331 ( 1995 ) .
m. sanchez [ soudan-2 collaboration ] , int .
j. mod .
a * 16s1b * , 727 ( 2001 ) .
k. daum [ frejus collaboration . ] , z. phys .
c * 66 * , 417 ( 1995 ) . c. berger _
[ frejus collaboration ] , phys .
b * 227 * , 489 ( 1989 ) .
m. aglietta
_ et al . _ [ the nusex collaboration ] , europhys .
lett . * 8 * , 611 ( 1989 )
. s. ragazzi _ et al .
_ , _ prepared for 9th moriond workshop : tests of fundamental laws ( particle physics , astrophysics , atomic physics ) , les arcs , france , 21 - 28 jan 1989_. m. ambrosio _ et al .
_ [ macro collaboration ] , phys .
b * 517 * , 59 ( 2001 ) [ arxiv : hep - ex/0106049 ] .
g. giacomelli and m. giorgini [ macro collaboration ] , arxiv : hep - ex/0110021 .
b. t. cleveland _
et al _ , astrophys .
j. * 496 * , 505 ( 1998 ) j. n. bahcall , m. h. pinsonneault and s. basu , astrophys .
j. * 555 * , 990 ( 2001 ) sage collaboration , j. n. abdurashitov _
et al _ , phys .
* c 60 * 055801 ( 1999 ) ; sage collaboration , v. gavrin , _ solar neutrino results from sage _ , in _ neutrino 2000 _ , proc .
of the xixth international conference on neutrino physics and astrophysics , nucl .
phys . * b91 * proc .
, 36 ( 2001 ) gallex collaboration , w. hampel _
et al _ , phys . lett .
* b 447 * , 127 ( 1999 ) gno collaboration , m. altmann _
et al _ , phys
* b 490 * , 16 ( 2000 ) ; gno collaboration , e. bellotti _
et al _ , _ first results from gno _ , in _ neutrino 2000 _ , proc .
of the xixth international conference on neutrino physics and astrophysics , nucl .
phys . * b91 * proc .
suppl . , 44 ( 2001 ) kamiokande collaboration , y. fukuda _ et al _ , phys .
lett . * 77 * , 1683 ( 1996 ) super - kamiokande collaboration , s. fukuda _ et al _ , phys .
lett . * 86 * , 5651 ( 2001 ) ; m. b. smy , arxiv : hep - ex/0202020 m. b. smy arxiv : hep - ex/0206016 .
sno collaboration , q. r. ahmad _ et al _ , phys .
lett . * 87 * , 071301 ( 2001 ) q. r. ahmad _ et al .
_ [ sno collaboration ] , arxiv : nucl - ex/0204008
. q. r. ahmad _ et al .
_ [ sno collaboration ] , arxiv : nucl - ex/0204009 .
p. aliani , v. antonelli , m. picariello and e. torrente - lujan , arxiv : hep - ph/0111418 , will appear on nucl .
* b * , issue 634 ( 1,2 ) ( 2002 ) p. aliani , v. antonelli , r. ferrari , m. picariello and e. torrente - lujan , arxiv : hep - ph/0205053 .
e. torrente - lujan , phys .
d * 59 * ( 1999 ) 093006 [ arxiv : hep - ph/9807371 ] .
e. torrente - lujan , phys .
d * 59 * ( 1999 ) 073001 [ arxiv : hep - ph/9807361 ] ; e. torrente - lujan , phys .
b * 441 * , 305 ( 1998 ) [ arxiv : hep - ph/9807426 ] ; v. b. semikoz and e. torrente - lujan , nucl .
b * 556 * , 353 ( 1999 ) [ arxiv : hep - ph/9809376 ] ; e. torrente - lujan , phys .
b * 494 * , 255 ( 2000 ) [ arxiv : hep - ph/9911458 ] .
j. n. bahcall , m. c. gonzalez - garcia and c. pena - garay , jhep * 0108 * , 014 ( 2001 ) [ arxiv : hep - ph/0106258 ] ; j. n. bahcall , m. c. gonzalez - garcia and c. pena - garay , jhep * 0204 * , 007 ( 2002 ) [ arxiv : hep - ph/0111150 ] ; a. y. smirnov , `` global analysis with sno : toward the solution of the solar neutrino problem,'',published in venice 2001 , neutrino oscillations , 43 - 64 ; g. l. fogli , e. lisi , d. montanino and a. palazzo , phys .
d * 64 * , 093007 ( 2001 ) [ arxiv : hep - ph/0106247 ] ; m. maris and s. t. petcov , phys .
b * 534 * , 17 ( 2002 ) [ arxiv : hep - ph/0201087 ] ; m. maltoni , t. schwetz and j. w. valle , phys .
d * 65 * , 093004 ( 2002 ) [ arxiv : hep - ph/0112103 ] ; a. bandyopadhyay , s. choubey , s. goswami and k. kar , phys .
b * 519 * , 83 ( 2001 ) [ arxiv : hep - ph/0106264 ] ; a. m. gago , m. m. guzzo , p. c. de holanda , h. nunokawa , o. l. peres , v. pleitez and r. zukanovich funchal , phys .
d * 65 * , 073012 ( 2002 ) [ arxiv : hep - ph/0112060 ] ; m. v. garzelli and c. giunti , phys .
d * 65 * , 093005 ( 2002 ) [ arxiv : hep - ph/0111254 ] .
p. aliani , v. antonelli , r. ferrari , m. picariello and e. torrente - lujan , arxiv : hep - ph/0205061 .
p. aliani , v. antonelli , m. picariello and e. torrente - lujan , nucl .
suppl . * 110 * , 361 ( 2002 ) [ arxiv : hep - ph/0112101 ] .
g. bellini , recent developments in the borexino project , talk given at _
neutrino 2002 _ , xxth international conference on neutrino physics and astrophysics , may 2002 , munich e. meroni , nucl . phys .
suppl . * 100 * , 42 ( 2001 ) ; g. ranucci _ et al . _
[ borexino collaboration ] , nucl .
suppl . * 91 * , 58 ( 2001 ) ; s. bonetti _ et al _ [ borexino coll . ] , nucl .
* 28 a * , 486 ( 1992 ) s. t. petcov and m. piai , phys .
b * 533 * , 94 ( 2002 ) [ arxiv : hep - ph/0112074 ] .
s. schonert , t. lasserre and l. oberauer , arxiv : hep - ex/0203013 j.h .
christenson _ et al _ , phys .
* 13 * , 138 ( 1964 ) .
j. n. bahcall , m. c. gonzalez - garcia and c. pena - garay , arxiv : hep - ph/0204314 .
v. barger , d. marfatia , k. whisnant and b. p. wood , arxiv : hep - ph/0204253 .
a. bandyopadhyay , s. choubey , s. goswami and d. p. roy , arxiv : hep - ph/0204286 ; r. foot and r. r. volkas , arxiv : hep - ph/0204265 ; s. pascoli and s. t. petcov , arxiv : hep - ph/0205022 .
p. c. de holanda and a. y. smirnov , arxiv : hep - ph/0205241 .
a. strumia and f. vissani , jhep * 0111 * , 048 ( 2001 ) [ arxiv : hep - ph/0109172 ] ; a. strumia , c. cattadori , n. ferrari and f. vissani , arxiv : hep - ph/0205261 . | we present a short review of the existing evidence in favor of neutrino mass and neutrino oscillations which come from different kinds of experiments .
we focus our attention in particular on solar neutrinos , presenting a global updated phenomenological analysis of all the available data and we comment on different possible future scenarios .
_ expanded version of the contribution to appear in the proceedings of `` les rencontres de physique de la vallee daoste '' , february 2002 .
_ # 1#2#3 ann .
phys .
( ny ) * # 1 * ( 19#2 ) # 3 # 1#2#3 ann .
rev .
nucl . part .
sci . * # 1 * ( 19#2 ) # 3 # 1#2#3 nucl .
phys .
* b # 1 * ( 19#2 ) # 3 # 1#2#3 nucl
. phys .
( proc .
suppl.)*b # 1 * ( 19#2 ) # 3 # 1#2#3 phys
. lett .
* b # 1 * ( 19#2 ) # 3 # 1#2#3 phys . rev . * d # 1 * ( 19#2 ) # 3 # 1#2#3 phys . rep .
* # 1 * ( 19#2 ) # 3 # 1#2#3 phys .
rev .
lett . *
# 1 * ( 19#2 ) # 3 # 1#2#3 prog .
. phys . *
# 1 * ( 19#2 ) # 3 # 1#2#3 rev .
mod .
phys . *
# 1 * ( 19#2 ) # 3 # 1#2#3 zeit .
fr physik * c # 1 * ( 19#2 ) # 3 # 1#2#3 mod .
. lett . * a # 1 * ( 19#2 ) # 3 # 1#2#3 sov .
j. nucl .
phys .
* # 1 * ( 19#2 ) # 3 # 1#2#3 yad . fiz . * # 1 * ( 19#2 ) # 3 # 1#2#3 nuovo cim .
* # 1 * ( 19#2 ) # 3 # 1#2#3 jetp lett . *
# 1 * ( 19#2 ) # 3 # 1#2#3 _ ibid . _
* # 1 * ( 19#2 ) # 3 cern - th-2002 - 213 + ifum-721/ft + ftuam-02 - 314 * the solar neutrino puzzle : present situation and future scenarios * + p. aliani@xmath0 , v. antonelli@xmath0 , r. ferrari@xmath0 , m. picariello@xmath0 , e. torrente - lujan@xmath1 + @xmath2 _ dip .
di fisica , univ .
di milano _ , _ and infn sez .
milano , via celoria 16 , milano , italy _
+ @xmath3 _ dept .
fisica teorica c - xi , univ .
autonoma de madrid , 28049 madrid , spain , _
+ @xmath4 _ cern th - division , ch-1202 geneve _
+ = 14.5pt |
the study of the cellular automata(ca ) has been one of the most important fields for the understanding of the complex systems . and
many efforts have been carried out for the understanding of it .
various patterns had been generated by choosing the transition rules , and the patterns have been classified by wolfram roughly into four classes@xcite .
langton has introduced parameter @xmath0 and showed that as @xmath0 is increased the pattern changes from class i(homogeneous),class ii(periodic ) to class iii(chaotic ) . and at some cases , the class iv(edge of chaos ) is realized between the two@xcite .
+ the @xmath0 parameter could classify the qualitative behavior of the ca , but it is not enough@xcite .
it is well known that at the same @xmath0 the time evolution of the configuration shows periodic pattern(class ii ) , chaotic one(class iii ) or edge of the chaos pattern(class iv ) depending on the initial seed of the random number generator .
the reason or mechanism for it has not been known yet ; we have no way to control the pattern classes at fixed @xmath0 . and the transition of periodic to chaotic pattern
is taken place at rather wide range of @xmath0 .
therefore it is a natural anticipation that more parameters are necessary to get the quantitative understanding of the pattern classes of the ca . in this article
we report another parameter which classifies the pattern classes more quantitatively . if the capacity to support computation is realized at the edge of the chaos@xcite , we have acquired a method to reach there more effectively . in the section 2
, we will briefly summarize our notations and explain the key discovery which leads us to the understanding of the structure of the rule table .
it strongly indicates that the rules which destroy the schema of the quiescent state play an important role to distinguish the pattern classes . in the section 3 , guided by the discovery of the section 2 , we classify the rule table according to the destruction and construction of the schema of the quiescent states .
the rule which destroys the longest schema of the quiescent state plays strongest effects for the patterns .
then we introduce a new parameter force for each rules which destroy the schema . with this parameter
, we determine the phase diagram in the @xmath0 and the force plane .
it clearly explains why at fixed @xmath0 the different patterns are generated and we could control the pattern classes more quantitatively .
section 4 is devoted to the discussions and the conclusions , where effects of the structure of the rule table will further be discussed .
in order to make the argument concrete , in this article we focus on the one dimensional cellular automata with 5-neighbors and 4-states .
but the qualitative results are independent of the details of the models .
this point will be discussed in section 4 .
we follow the definitions of langton@xcite but for the self consistentness we will briefly summarize them . in our study
, the sites consist of 150 cells with the periodic boundary condition , which are denoted as @xmath2 .
the @xmath3 represents the time step which take integer value , and the @xmath4 is the position of site which range from @xmath5 to @xmath6 .
the @xmath2 takes the value @xmath7 and @xmath8 , and the state @xmath5 is taken to be quiescent state@xcite .
the set of the states @xmath2 at fixed @xmath3 is called the configuration at time @xmath3 . the configuration at time @xmath9 is determined by that of time @xmath3 by the following local relation , @xmath10 the set of the mappings @xmath11 is called rule table .
the rule table consists of @xmath12 mappings , which are selected from the totally @xmath13 elements .
the @xmath0 parameter is defined as follows@xcite , @xmath14 where @xmath15 is the number that @xmath16 in eq.[r_table ] is not equal to @xmath5 . in other word
the @xmath0 is a probability that the rules do not select the quiescent state in next time step . in the following we set the rule tables randomly according to the probability @xmath0@xcite , and
the initial configurations are also set randomly .
the time history of the configurations is called pattern .
the patterns are classified by wolfram@xcite roughly into four classes .
it has been known that as the @xmath0 increases the most frequently generated patterns change from homogeneous(class i ) , periodic(class ii ) to chaotic(class iii ) and at the region between class ii and class iii , the edge of chaos is located(class iv ) . in order to find the reason why the different patterns are generated at the same @xmath0
, we had started to collect a lot of rule tables with different pattern classes . and
by studying them , we had tried to find the differences between rule tables .
we had fixed @xmath18 , which correspond to @xmath17 .
we had chosen this @xmath0 because at this point the chaotic , edge of chaos and periodic patterns are generated in the similar ratio . by changing the random number ,
we have gathered a few tens of the rule tables and classified them into chaotic , edge of chaos and periodic ones .
first we studied whether the pattern classes are sensitive to the initial configurations or not .
we fixed the rule table and changed the initial configurations .
the details of the patterns depend on the initial configurations but the pattern classes are not changed@xcite .
then the difference of the pattern classes is due to that of the rule tables and our target is addressed to the study of them .
after some trial and error , we have found the fact that for rule tables which generate the edge of the chaos and periodic patterns , the transition table @xmath19 was almost always equal to @xmath5 , while for that of the chaotic pattern , it was not equal to @xmath5 in majority of cases . for about 10 rule tables of each pattern classes , we get following results for the probability that the rule tables have @xmath20 .
@xmath21 for periodic plus edge of chaos rule tables , + @xmath22 for chaotic rule tables . in this article
, we call the pattern as edge of chaos when its transient length@xcite is longer than @xmath23 time steps .
these difference given by eq.[perio ] and eq.[chaos ] has lead us the following idea .
let us think of the schemas of quiescent state @xmath5 with length longer than five . if the rule @xmath24 is included in the rule table , they are necessarily broken .
then it pushes the pattern to that of chaos .
if our idea is correct , we could artificially change the rule tables of the chaos to that of periodic by replacing @xmath25 to @xmath20 .
we have tried the replacements for about 10 rule tables . in all cases
we have succeeded to change the rule tables from chaotic to periodic or periodic to chaotic by a replacements of the @xmath19 .
an example is shown in the figure [ pattern_change ] .
the change in the pattern is dramatical but the differences in the rule table are only two , which has been quite exciting for us .
this discovery was a key hint to lead us to the hypothesis that the rules which breaks the schema of the quiescent state will determine the pattern classes .
in order to test the hypothesis of the previous section , we group the rule table into the four types according to the operation on the schema of quiescent state . in the following , greek characters in the rule table represent @xmath7 and @xmath8 while roman ones @xmath26 and @xmath8 .
type 1 : @xmath27 .
+ the rules in this type destroy the schemas of quiescent state .
type 2 : @xmath28 .
+ the rules of this type conserve them . +
type 3 : @xmath29 . + the rules of this type develop the schema . + type 4 : @xmath30 . + the rules in this type do not affect the schema in next time step .
the sum of the number of the type 1 and type 2 rules is 256 , while that of type 3 and type 4 rules is 768 .
how many numbers of each type of rules are included in the rule table is determined randomly according to the probability @xmath0 .
the type 1 rules are further classified into five groups according to the length of the schema which breaks .
they are shown in the table [ destruction ] .
the classification of the rules in type 1 .
[ destruction ] [ cols="^,^,^,^,^,^,^ " , ] the results in table [ relative_force ] are natural in the following sense
. if six d4 rules are present for the rule table , then the presence of d5 has no effects , because the length 5 schema of quiescent state could not be made .
they have roughly similar effects as one d5 rule , then the strength of the d4 will approximately be equal to @xmath31 of that of d5 .
similarly the strength of the d3 and d2 will be about @xmath32 and @xmath33 respectively .
the results given in the table [ relative_force ] are not very different from these naive expectations . with these results for the coefficients of eq.[force_sum ] , we determine the transition points by means of the force .
we call it as critical force(@xmath34 ) .
the results are shown in the fig.[phase_diagram ] . in the fig.[phase_diagram ] , we have added @xmath35 calculated from the naive expectations for the relative strengths . in this case too , errors are estimated by the jackknife method .
the difference between them are not large .
this figure distinguishes the pattern classes more quantitatively , and visualizes why at the fixed @xmath0 the three patterns chaos , edge of the chaos and periodic coexist .
the impressive point is that the transition occurs about @xmath36 .
this means that in many cases the replacements of d5 will change the pattern classes , which is consistent with the experiments at section 2 .
we should like to notice that the phase diagram shown in fig .
[ phase_diagram ] is an average over about 60 to 85 sets of the critical combinations , which belong to different rule tables .
@xmath37 difference of patterns with @xmath25 and @xmath20 + by carrying out the replacements rp4s and rp3s , we could artificially obtain the rule table of periodic pattern with @xmath25 . in this case
the length of the schema of quiescent state could not develop longer than 4 , then the patterns will have thin stripes in the time direction .
while for the patterns generated by the rule table without d5 , there is very small probability that the thin stripes are created . then there will be clear difference between the two patterns .
an example is shown in the fig.[periopattern ] .
the differences in patterns is again dramatical while those in rule table are only 2 .
similarly for the patterns of an edge of the chaos , when the rule tables have d5 , the patterns are drawn on the background of the thin stripes as shown in the fig.[periopattern]a , while if they do not include d5 , patterns are drawn essentially on the quiescent states .
the differences between these two patterns are impressive .
similar situation is observed for the pattern when six d4s are included in the rule table . in this case , the maximum length of the schema of quiescent state is 3 , that the width of the stripe becomes more narrow .
the existence of the rules of the table [ destruction ] has strong effects for the details of the patterns .
@xmath37 fluctuation of rule tables around @xmath38 + when the rule tables are generated randomly , how many type 1 rules are included , is determined randomly by the probability @xmath0 .
the largest fluctuation is realized for d5 , because it has only one element .
next is d4 and so on . and the transition is taken place about @xmath39 to @xmath40 of the force expressed in the unit of @xmath41
. then the change of the pattern classes are easily realized by the fluctuation of @xmath42 and @xmath43 at fixed @xmath0 .
this fact explains why the transition and edge of chaos is scattered rather wide range in @xmath0 .
also the fluctuation is largest at @xmath44 .
then the coexistence of the periodic pattern and chaotic pattern is most likely to be realized around this region .
as the edge of chaos is located at the boundary of these two pattern classes , it is also realized most frequently around @xmath45 .
especially when the number of the states in cell is equal to two .
@xmath37 edge of chaos and order of transition + when the replacements of rp3 or rp2 are carried out , we find the edge of chaos(very long transient lengths ) in many cases . sometimes they are observed rather wide range in @xmath46 or @xmath47 .
this seems to indicate that in many cases , the transition is second order like .
but the widths in the ranges of @xmath46 or @xmath47 are different from each other and there are cases where the widths are less than one unit in the replacement of d2 rules(first order like ) .
what is the origin of the difference in the width is an open problem and may be studied by taking into account the effects of the type 3 rules .
it is a very interesting problem under what condition , the transition becomes first order like or second order like .
@xmath37 special cases of the rule table + the relative strength of the force given by the table [ relative_force ] could not be applied for some special cases .
when the rule table contains all the 33 d3 rules , the schemas of the quiescent state with length 4 could not develop , then the presence of the d4 have no chance to work . in this case , the force from the d4 and d5 are equal to zero .
the results given in the table [ relative_force ] do not include in these special cases .
but these special rule tables are very hard to be realized as far as they are determined randomly .
the results for the table [ relative_force ] are the average over the rule tables which are typically obtained randomly according to the probability @xmath0 .
we have made a few replacements experiments at more extreme @xmath0 . at @xmath48 , in most cases , a randomly generated rule tables belong to chaotic pattern .
we have succeeded to change them to the rule table of periodic pattern by the replacements of rp5 , rp4s and rp3s . on the other hand at @xmath49 ,
most of the generated rule tables were periodic ones . in this case too , the replacements rc5 and a few rc4s make the rule table to that of chaotic pattern classes . then the conclusion that the number and combination of the type 1 rule determine the pattern classes is correct for more wider space of rule tables .
however when @xmath0 exceed @xmath50 , the state @xmath5 lost its role as a quiescent state .
then the analysis in this article will not be applied there .
@xmath37 other one dimensional cellular automata + the similar study has been carried out for the one dimensional 3-neighbors and 16-states ca . in this case
there are 3 groups of rules which destroy the schemas of quiescent state ; @xmath51 and @xmath52 , @xmath53 and @xmath54 .
the results are very preliminary , but we have succeeded to change the chaotic patterns to periodic one by making only one set of replacements at @xmath55 . @xmath56 it needs some more studies to obtain the quantitative results .
but the result that the rules which breaks the schemas of quiescent state play main role for the determination of the pattern classes and the qualitative phase diagram will be common for all ca including two dimensional ones .
@xmath37 conclusions + the pattern in the ca is strongly affected by the type 1 rules , which are classified in the table [ destruction ] . around @xmath57 region ,
the replacements of d5 and a few d4s change the pattern classes .
the number of the replacements has been always less than @xmath58 if rp5 and rp4 or rc5 and rc4 are included .
the type 1 rules have a hierarchy in their effect to push the rule table toward chaotic one as shown in eq .
[ order_force ] .
this property is studied more quantitatively by introducing force for the rules in table [ destruction ] .
then we have obtained the phase diagram of ca .
it is determined by many approximations and assumptions .
but it tell us how to control the pattern classes at fixed @xmath0 .
the edge of the chaos of the ca is realized as the result of the delicate balance between the creation of the schemas of quiescent state by type 3 rules and the destruction of them by the type 1 rules . by making replacements defined by the eq.[toperio ] or eq .
[ tochaos ] , we could control the pattern classes more efficiently than using only the parameter @xmath0 .
the results obtained in this article is a beginning of the more quantitative study of the rule tables and pattern classes .
the analysis of the structure of the rule table including the effects of the type 3 rules will be necessary to understand the order of the transition and the detailed properties of the edge of chaos , which will be reported elsewhere .
9 s. wolfram , physica d10(1984 ) 1 - 35 c.g.langton , physica d42(1990 ) 12 - 37 w.li , n.h.packard and c.g.langton , physica d45(1990 ) 77 - 94 m. mitchell , j.p.crutchfield and p.t .
hraber , santa fe institute studies in the science of complexity , proceedings volume 19 .
reading , ma : , addison - wesley .
online paper , http://www.santafe.edu/ mm/ paper-abstracts.html#dyn-comp-edge t. suzudo , crystallisation of two - dimensional cellular automata , complexity international , vol .
on line journal , + http://www.csu.edu.au/ci/vol06/suzudo/suzudo.html .
see appendix . | in addition to the @xmath0 parameter , we have found another parameter which characterize the class iii , class ii and class iv patterns more quantitatively . it explains why the different classes of patterns coexist at the same @xmath0 . with this parameter
, the phase diagram for an one dimensional cellular automata is obtained .
our result explains why the edge of chaos(class iv ) is scattered rather wide range in @xmath0 around @xmath1 , and presents an effective way to control the pattern classes . |
clusters of galaxies are the largest virialized objects in the present universe . according to the standard scenario of structure formation in the universe , larger objects form more recently .
thus , clusters of galaxies are the virialized objects which form in the most recent epoch in the universe .
in fact , some clusters are forming now , which is evident from the moving substructures that are found through x - ray observations ( e.g. markevitch et al .
, 2000 , 2002 ; vikhlinin et al . ,
2001 ) .
a moving substructure causes various characteristic structures in the intracluster medium ( icm ) .
a bow shock and a contact discontinuity will form in front of it .
the latter most likely corresponds to a `` cold front '' which is found by _
chandra _ in a number of clusters ( e.g. markevitch et al . , 2000 ; vikhlinin et al . , 2001 ) .
moving substructures generate turbulence in the icm through fluid instabilities @xcite .
turbulence probably plays an important role in cluster evolution
. it may have a significant impact on the transport and mixture of heavy elements and on those of thermal energy .
fluid turbulence probably generates magnetic turbulence , which accelerates non - thermal particles and causes various high energy phenomena in the intracluster space ( e.g. ohno et al . , 2002 ; fujita et al . , 2003 ; brunetti et al . ,
the _ astro - e2 _ satellite will detect broadened lines due to turbulent motion @xcite . in order to study the above - mentioned issues ,
it is crucial to clarify the generation processes and structure of the icm turbulence .
in the present study , we used the roe tvd scheme to follow the dynamical evolution of the icm ( see hirsch , 1990 ) .
the roe scheme is a godunov - type method and is based on a linearized riemann solver @xcite .
it is relatively simple and good at capturing shocks without any artificial viscosity . using the muscle approach and a minmod tvd limiter
, we obtained second - order accuracy without any numerical oscillations around discontinuities . to avoid negative pressure
, we solved the equations for the total energy and entropy conservation simultaneously .
this method is often used in astrophysical hydrodynamical simulations where high mach number flows can occur @xcite .
our initial conditions are as follows .
we set a subcluster in the center of the simulation box .
the subcluster s gravitational potential is represented by that of a king distribution , whose total mass and core radius are @xmath0 and @xmath1 kpc , respectively .
the subcluster s radius is @xmath2 kpc . inside the subcluster
, the gas is assumed to be in hydrostatic equilibrium with an isothermal temperature distribution .
we set the initial temperature inside the subcluster so that @xmath3 .
we assume that the mass density of the gas is a tenth of that of the dm in the center . outside the subcluster
, the gas pressure is the same as that on the outer boundary of the subcluster and the density is half of the density on that boundary . only in front of the subcluster we set the initial velocity to twice the sound speed .
the size of the simulation box and the number of grid points are ( 2 mpc)@xmath4 and @xmath5 , respectively .
we set constant inflow boundary conditions only for the front boundary .
free boundary conditions are adopted for the other boundaries .
figure [ fig : density ] shows snapshots of the density distribution on the @xmath6 surface at @xmath7 and 3 gyr .
two shocks propagate forward and backward .
the shock which propagates backward has left the simulation box before @xmath8 gyr .
thus , only the forward shock is seen in figure [ fig : density ] .
the substructure s gas in the outer region is stripped off behind the shock because of the ram pressure .
in addition , a kelvin - helmholtz instability developed on the boundary between the substructure and the ambient gas , and prominent eddy - like structures form just behind the substructure . surface at @xmath7 and @xmath9 gyr .
the substructure s gas in the outer region is stripped off behind the shock because of the ram pressure . a kelvin - helmholtz instability developed on the boundary between the substructure and the ambient gas , and prominent eddy - like structures form just behind the substructure . ] although the eddy - like structures are clearly seen in figure [ fig : density ] , in actual observations we can get only information integrated along the line - of - sight . therefore , these structures may be less clear in observed quantities .
figure [ fig : xrsb ] shows snapshots of the x - ray surface distribution seen along the @xmath10-axis .
we assume that the volume emissivity is proportional to @xmath11 , where @xmath12 and @xmath13 are the gas density and temperature , respectively . because of contamination of the foreground and background components , the eddy - like structures around the subcluster become less clear . and @xmath9 gyr
. the line - of - sight is just along the @xmath10-axis .
eddy - like structures around the subcluster become less clear because of contamination of the foreground and background components . ]
the _ astro - e2 _ satellite has a high resolution x - ray spectrometer xrs
. it will be able to detect @xmath14 km s@xmath15 gas flows through doppler broadening and/or shift of the emission lines . because of the symmetry , no line shift is expected in our simulations when the line - of - sight is along the @xmath10-axis . on the other hand ,
line broadening is expected because of the bulk flows and turbulent motion .
the line - of - sight velocity dispersion ( or standard deviation , equivalently ) is a good estimator of line - broadening .
figure [ fig : rms_vz ] shows the emissivity - weighted standard deviation distribution of the line - of - sight velocity component . around the contact discontinuity in front of the substructure ,
the root mean square velocity is @xmath16 km s@xmath15 because of the bulk flows along the convex boundary . behind the substructure ,
on the other hand , it becomes @xmath17 km s@xmath15 because of the turbulent motion associated with the eddies . and @xmath9 gyr .
around the contact discontinuity in front of the substructure , the root mean square velocity is @xmath16 km s@xmath15 because of the bulk flows along the convex boundary . behind the substructure , on the other hand , it becomes @xmath17 km s@xmath15 because of the turbulent motion associated with the eddies . ]
we investigate the gas motion in and around a moving substructure in a cluster of galaxies using a three dimensional roe - tvd hydrodynamic code . after the icm in the outer region of the substructure
is stripped off because of the ram pressure , a kelvin - helmholtz instability develops on the boundary between the substructure and the ambient icm .
then , eddy - like structures form there and the cold gas clouds originating from the substructure fly backwards .
the line - of - sight velocity dispersion of the icm becomes large around and in the back of the substructure remnant when we see the system from the direction perpendicular to the substructure motion .
the root mean square velocity becomes @xmath17 km s@xmath15 , which is large enough to be detected by the _ astro - e2 _ xrs .
although in this paper we only showed the results in a rather simple case where a subcluster is moving at a constant velocity in a uniform background , we are planning to investigate more realistic situations where the ambient icm properties can vary with time .
for instance , simulations of a subcluster infalling into a larger cluster and sloshing near the cluster center will be useful . in this work
, we neglected the magnetic field in the icm .
however , it is possible that the magnetic field plays a crucial role to suppress the development of kelvin - helmholtz instabilities @xcite . on the other hand ,
the evolution and structure of mhd turbulence contains important information about particle acceleration in the icm .
therefore , high resolution mhd simulations will be important as future work .
the numerical computations were carried out on the vpp5000 at the astronomical data analysis center of the national astronomical observatory , japan , which is an inter - university research institute of astronomy operated by the ministry of education , science , culture , and sports .
m. t. was supported in part by a grant - in - aid from the ministry of education , science , sports , and culture of japan ( 16740105 ) . | clusters of galaxies form through major merger and/or absorption of smaller groups .
in fact , some characteristic structures such as cold fronts , which are likely relevant to moving substructures , are found by _
chandra_. it is expected that moving substructures generate turbulence in the intracluster medium ( icm ) .
such turbulence probably plays a crucial role in mixture and transport of gas energy and heavy elements , and particle acceleration .
the _ astro - e2 _ satellite , which is planned to be launched in 2005 , will detect broadened lines due to turbulent motion . in order to explore the above - mentioned issues , it is important to investigate the generation processes and structure of icm turbulence .
we investigate the icm dynamical evolution in and around a moving substructure with three - dimensional hydrodynamical simulations .
eddy - like structures develop near the boundary between the substructure and the ambient icm through kelvin - helmholtz instabilities .
because of these structures , characteristic patterns appear in the line - of - sight velocity distribution of the icm . clusters of galaxies , intracluster medium , turbulence |
the study of transport properties in a system of charged fermions is an interesting subject in areas as different as physics of plasma , quantum chromodynamics or metals . in particular , the measure of
the electrical conductivity is a very difficult and yet interesting problem , specially in presence of nonperturbative effects .
in such a case , numerical methods are called for . in this work ,
we want to show that lattice - regularized euclidean field theories can be useful in this respect , at least in the limit of vanishing temperature .
however , it should be emphasized that the so called _ sign - problem _ needs still to be overcome in many cases @xcite ( see , though , refs .
@xcite for some successful simulations at finite density ) . in this paper
, we restrict ourselves to a model consisting of fermions that only interact with an external electromagnetic field . in spite of its simplicity
, it shares many properties with more realistic models , and it can be considered as a necessary first step to check any numerical method which could be used in models with dynamical interactions .
we consider the standard @xmath0 lattice - action , with wilson fermions and finite chemical potential , but with the gauge variables held fixed .
we shall study the residue of the pole of the electrical conductivity at zero - frequency , which is purely imaginary .
since a non - vanishing value for this residue unambiguously signals a conducting phase , this is a rather interesting quantity in our opinion . in order to obtain it ,
we measure the electrical - current induced in the system by an external electric field .
this technique requires a numerical calculation even in the case of an external spatially - homogeneous time - dependent electromagnetic field .
the delicate point , however , is that our electric field varies in euclidean time .
one can nevertheless assume that there is a linear relation between the euclidean current and the euclidean electric field , at least for small fields .
this _ euclidean conductivity _ presents a pole whose residue can be straightforwardly measured . to check that the obtained result is physical , we follow a very elegant procedure due to kohn@xcite .
he showed that the _ real - time _ residue can be measured by studying the sensitivity of the ground - state energy to an external aharonov - bohm electromagnetic field .
we show how can this be done in the lattice formalism , and , in this particularly simple case of free fermions , we calculate it ( unfortunately , the kohn recipe seems really hard to use in a monte carlo study of a self - interacting problem ) .
although at present we lack a rigorous proof of the equivalence of both calculations , its excellent numerical agreement gives a strong support to the linear response method .
let us consider a model of wilson fermions@xcite in a lattice of spacing @xmath1 in the three spatial directions and @xmath2 in the temporal one , coupled to an external electromagnetic field .
we denote by @xmath3 to the quotient @xmath4 .
the partition function can be written as ( the @xmath5 superscript stands for complex conjugation ) @xcite @xmath6=\int\prod_z\,{\mathrm d}\psi_z { \mathrm d}\overline{\psi}_z \exp\left[\sum_{x , y}\overline{\psi}_x m_{xy}(u)\psi_{y}\right]\label{pathint}\,,\ ] ] @xmath7 + [ ( 2m+6r_\mathrm{s})\lambda+2r_\mathrm{t}]\delta_{x , y}\ , \nonumber\end{aligned}\ ] ] where @xmath8 , @xmath9 being the gauge field , and @xmath10 , @xmath11 are the anticonmuting grassmann fermionic fields .
the indices @xmath12 run on the points of the four - dimensional space - time lattice .
we impose periodic boundary conditions for the gauge field , and periodic in space but antiperiodic in time ( @xmath13 ) for the grassmann field .
the site @xmath14 is the neighbor of @xmath15 in the @xmath16 direction . for finite temporal length , @xmath17 ,
the system is at finite temperature @xmath18 . in this paper
we will only consider the zero temperature ( @xmath19 ) limit .
we follow the prescription of introducing the chemical potential through an imaginary gauge field @xmath20 @xcite , which is fairly convenient for analytical calculations .
the wilson parameter , @xmath21 , can be taken different for the spatial and time directions . in the limit @xmath22 with @xmath1
fixed the model describes a spatial lattice with a continuous time ( as electrons in a metal ) , while for a continuum field theory both spatial and time continuum limits should be taken .
we shall use the following representation for the ( euclidean ) gamma matrices @xmath23 where @xmath24 are the pauli matrices . to define the electric four - current in the lattice we recall that in the space - time continuum limit it is defined as @xmath25 that can be obtained as a logarithmic derivative of the partition function respect to the gauge - field .
this calculation can be exactly mimicked on the lattice noticing that a change in the link variable should be of the form @xmath26 . in this way one obtains @xcite : @xmath27 where now @xmath28 the @xmath29 component is just the electric charge density that one encounters by differentiating with respect to @xmath30 the free energy density @xcite . moreover , from the gauge invariance of the determinant of the fermionic matrix , @xmath31 , it is straightforward to prove the lattice continuity equation , for any configuration of the electromagnetic - field : @xmath32 eqs .
( [ j0 ] ) and ( [ ji ] ) can be written free of grassmann variables as @xmath33+{\mathrm e}^{-\lambda\mu } u^*_{x,0 } \mathrm{tr } [ ( \gamma_0+r_\mathrm{t } ) m^{-1}_{x , x+\hat 0}]\ , \label{j0ng}\\ -\langle j_{x , i}\rangle&=&\lambda u_{x , i } \mathrm{tr } [ ( \gamma_i - r_\mathrm{s } ) m^{-1}_{x+\hat \imath , x } ] + \lambda u^*_{x , i } \mathrm{tr } [ ( \gamma_i+r_\mathrm{s } ) m^{-1}_{x , x+\hat \imath}]\ , \end{aligned}\ ] ] where tr stands for the trace over dirac indices . the above expressions and the relation @xmath34 allow to prove that @xmath35 in an uniform electrical field , the charge density should remain constant under field inversion , while the electrical current should change sign . therefore , from eq .
( [ jconj ] ) one expects the former to be real and the latter to be imaginary ( euclidean space - time ! ) . in absence of external fields ( @xmath36 )
the matrix @xmath31 can be diagonalized in fourier space , which allows to explicitly perform the functional integrals , and to compute the free energy and the propagator . for brevity
, we only quote the result for the charge density in the case @xmath37 , @xmath38 , that in the infinite volume limit reads ( see refs .
@xcite for similar calculations ) , @xmath39 where @xmath40\,,\\ \sigma(\mbox{\boldmath $ k$})&= & m+\sum_j\,(1-\cos k_j)\ .\label{sigma } \label{er1}\end{aligned}\ ] ] an useful quantity is the mechanical compressibility that , at zero - temperature coincides with the density of states : @xmath41 the density of states of the system present a typical band structure ( see the upper part of fig .
[ den - res ] , dashed line ) .
the upper limit of the band corresponds to the saturation due to the fermi statistics ( one particle per lattice - site ) . since the function @xmath42 is periodic , its gradient has zeroes in the brillouin zone , producing non - analiticies as the cusps in fig .
[ den - res ] ( the so - called van - hove singularities @xcite ) .
in a classical paper , kohn @xcite developed an elegant characterization of a conductor , at zero temperature .
his method allows the measurement of the following limit for the imaginary part of the electrical conductivity , @xmath43 , @xmath44 if this limit turns out to be non zero , the system is a conductor .
the construction is as follows .
the system of interest is constrained to verify periodic boundary conditions in the ( say ) first spatial direction , and immersed in an aharonov - bohm like electromagnetic field @xmath45 . with this choice of boundary conditions
the product @xmath46 is gauge invariant since it represents the magnetic flux traversing the system .
it can be shown that @xmath47 where @xmath48 is the ground - state energy and @xmath49 is the spatial volume .
it is crucial that the infinite limit volume is taken _ after _
the @xmath50 derivative is performed , since the effect of the aharonov - bohm field can be thought of as a change in the boundary conditions ( see below ) . in the infinite volume limit ,
the energy no longer depends on @xmath50 . in our case ,
as the free energy and the ground - state energy coincide in the zero temperature limit , we can study the residue in the following way @xmath51 where @xmath52 is obtained from the intensive free energy @xmath53 after subtracting the vacuum contribution : @xmath54 .
let us sketch the calculation .
the free energy should be calculated in a finite volume and at finite temperature .
we introduce our system in the aharonov - bohm electromagnetic field : @xmath55 this field can be transformed into a boundary effect by performing the following gauge transformation : @xmath56 so that @xmath57 excepting @xmath58 by direct inspection of the fermion matrix in eq . ( [ matriz ] ) , one can easily recognize that a system verifying periodic boundary conditions in the 1 direction in the field @xmath59 is equivalent the same system with no field at all , but verifying @xmath60 this amounts to substituting @xmath61 by @xmath62 in the momentum - quantification in a finite lattice . for a system of free fermions
the free energy can be now straightforwardly calculated .
once the @xmath50 derivative is performed , the zero temperature limit can be taken by transforming the @xmath63 sum into an integral .
we get , in the simplest case @xmath37 , @xmath38 , @xmath64 notice that for the empty system , @xmath65 , the integral vanishes , as well as for the full band @xmath66 , since @xmath67 is a periodic function of @xmath61 .
the three dimensional integrals ( [ condkohn ] ) can be performed using a monte carlo method .
the results are shown in fig .
[ den - res ] ( dashed line in the lower part ) .
in this section , we are going to reproduce the results of the sections [ model ] and [ elect ] by directly considering the integration of the partition function .
this method has the advantage of being generalizable to inhomogeneous external fields , and also when interacting dynamical fields are present .
examples of how to introduce an external field on an interacting lattice - gauge system can be found in refs .
@xcite . to compute the partition function
it is necessary to work in finite lattices , consequently , an infinite volume limit should be taken .
we have carried out measures in symmetric lattices of sizes @xmath68 and @xmath69 , with @xmath70 and @xmath71 .
for the hopping term , we have taken @xmath72 .
as the integral in the fermionic fields is gaussian , the computation of the electric current just requires the inversion of a @xmath73 matrix , @xmath74 being the space - time volume . the fermion matrix ( [ matriz ] ) being sparse , we have used a conjugate - gradient algorithm for the numerical inversion .
we first consider the density of states in a vanishing external field . in order to measure @xmath75
we invert the matrix at @xmath76 for @xmath77 small enough . in interacting systems
the derivative can be calculated in terms of connected correlation - functions @xcite .
the numerical results are plotted in fig .
[ den - res ] , upper part , together with the infinite volume values obtained analytically .
although the finite size effects are non negligible even in the larger lattices for most values of @xmath78 , there is a clear trend to the asymptotic values .
unfortunately , for an interacting system it is not immediate how to implement kohn s method for calculating the residue of the conductivity .
in fact , the free energy is rather hard to calculate with a monte carlo simulation and what one directly obtains are mean - values .
we are now going to present a different way of computing the residue , by directly measuring the system response to an external electrical field .
notice that the presence of an electric field requires a non - homogeneous vector potential and consequently the inversion of the fermion matrix can no longer be performed in closed analytical form .
this new recipe can be straightforwardly generalized to interacting systems , but its equivalence with the kohn s method is just an ansatz .
nevertheless the agreement is excellent , as we will show . by analogy with continuum electrodynamics , we want to study the electric current induced in the system by an external weak uniform electric field in the @xmath79 direction .
the conductivity ( in the frequency domain ) , will be the proportionality constant between the electrical current and the external field .
there are some subtleties that need to be considered when putting an external electric field on the lattice .
we take the gauge - field configuration ( @xmath80 ) @xmath81 @xmath82 notice that the quantization of the electric - field is due to the spatial boundary conditions . to preserve the translational symmetry ,
the displaced gauge field @xmath83 should be a gauge - transform of the one in eq .
( [ uelectrico ] ) . since the needed gauge transformation is analogous to eq .
( [ cambiogauge ] ) , it is easy to check that the condition that allows this transformation is the trivialness of the polyakov loop : @xmath84 with @xmath85 integer .
this condition also allows to transform the gauge field to the coulomb gauge @xmath86 .
if condition ( [ polyakov ] ) is violated , the translational invariance is lost and the electric current is no longer spatially homogeneous even on a homogeneous electric - field . however , with the correct field choice ( [ polyakov ] ) , we get a homogeneous electrical current aligned with the external electrical field , and imaginary as anticipated in eq .
( [ jconj ] ) . in order to directly compare with the results obtained with eq .
( [ condkohn ] ) , let us define @xmath87 if we want to stay within linear - response theory , we have to postulate a linear relation between the fourier transform of the electrical current @xmath88 and the external electrical field @xmath89 : @xmath90 notice that both @xmath88 and @xmath89 being real , @xmath91 .
however , the results can be more cleanly cast in terms of a modified fourier transform for the electrical field : @xmath92 the rationale for this is that the electrical field @xmath89 on the lattice lives mid - way between sites at times @xmath93 and @xmath94 .
the modified conductivity @xmath95 is related with the previous one by @xmath96 the nice feature of @xmath97 is that it turns out to be purely imaginary . in fig .
[ curvaconductividad ] we plot the imaginary part of @xmath97 as obtained from a field with @xmath98 and @xmath99 , ( from now on we shall only indicate the non - vanishing @xmath100 s ) , in a system of wilson fermions with @xmath70 , @xmath101 and @xmath102 , that is within the band energy - range and therefore with a non - vanishing fermi surface .
we see that for large frequencies the thermodynamic limit is reached in rather small lattices . however , at the minimal reachable frequency ( @xmath103 ) the conductivity is rapidly growing suggesting a singularity at zero frequency .
in fact , for a ( classical ) system of free particles of density @xmath85 we expect that @xmath104 will behave as @xmath105 notice that if @xmath104 has a pole at @xmath106 with a purely imaginary residue the same will hold true for @xmath97 , and both residues will be equal .
although the _ euclidean _ conductivity @xmath97 do not match the real - time one ( being imaginary , it can not fulfill the kramers - kronig relations ) , one can formally expect the residues to coincide in the passage from @xmath107 to @xmath108 .
this suggest to define the following quantity which will be the basic object of our study : @xmath109 in the @xmath110 limit , @xmath111 tend to the residue of the pole . in order to measure this ,
we have considered the smallest of possible external disturbances : @xmath112 .
our result is shown in fig .
[ den - res ] , lower part .
we see that the _ euclidean _ residue follows quite closely kohn s result , which in fact can be considered as the infinite volume limit for our calculation .
moreover , the physical picture is rather transparent : when the band is full , the system gets almost inert , while when the band is empty , it can be excited by the external field creating a hole in the dirac sea . since the smallest possible
excitation has frequency @xmath103 , to be compared with a gap @xmath113 , it is reasonable that at @xmath114 , the larger is the space - time lattice , the smaller is the system response .
in fact , notice that in fig .
[ den - res ] when @xmath78 is below the lower band limit , the curves get horizontal : in this range of @xmath78 the system can be only excited by crossing the gap between the dirac sea and the conduction band . and the gap is , of course , @xmath78 independent in a non - interacting system .
we remark that our results have been found within the linear response approximation .
we can control this approximation in several ways .
one is to study the fourier transform of the current , for frequencies at which the fourier transform of the external - field vanishes . in fig .
[ nolineal ] we show the zero - mode of the electrical current for the electrical - field @xmath115 .
we see that this non - linear effect tends to zero with growing lattice - size , which is quite reasonable since the minimum possible electric field is @xmath116 .
the non - linear corrections are oscillating , but modulated by a rapidly decaying function . roughly speaking , for the largest lattice the non linear effects are of the same order as the distance to the thermodynamic limit .
a further check can be done by comparing the residue obtained from the data in fig .
[ curvaconductividad ] with the one in fig .
[ den - res ] : in the @xmath117 lattice , the differences are at the @xmath118 level , while in the @xmath119 lattice the differences are at the @xmath120 level .
therefore , we believe that non - linear effects are under control for the not too small fields that we can deal with .
we have presented a simple way of studying the electrical conductivity of a system of wilson fermions at finite density and zero - temperature in a path - integral formalism . in particular , we have computed the residue of the zero frequency pole of the conductivity , by numerically considering the linear response to an external electric field , varying in euclidean time .
the results have been contrasted with an analytical computation based on a method proposed by kohn , and an excellent agreement has been found . as a further cross - check ,
we have computed the density of states both analytically and numerically in a finite lattice , obtaining a nice thermodynamic limit convergence .
it should be emphasized that in contrast with the analytical calculation which can only be done for a non interacting system ( or , at most , for simple external fields ) , the numerical calculations are easily generalizable to more complex models , as fermions self - coupled with quartic interactions or via a dynamic bosonic field .
an open , very interesting question is the possibility of extracting the full real - time electrical conductivity function from its _ euclidean _ counterpart .
we have shown that the residue of the zero - frequency pole can indeed be obtained
. it would be also very interesting to extend this approach to systems at finite - temperature .
this work was triggered during a discussion with f. guinea to whom we are indebted for many suggestions , discussions and bibliographical help .
the numerical calculations have being carried - out in the rtnn machines at the universities of zaragoza and complutense de madrid .
this work has been partially supported by cicyt , contracts aen97 - 1680,1693,1708 .
f. karsch , m - p .
lombardo ( editors ) , _ qcd at finite barion density _ , nucl
a642*(1998 ) proc .
supp . 1 - 2
. f. karsch , j.b .
kogut , h.w .
wyld , nucl .
phys . * b280*[fs18](1987)289 ; s.j .
hands , a. koci , j.b .
kogut , nucl . phys .
* b390*(1993)355 ; s.j .
hands , s. kim , j.b .
kogut , nucl . phys . * b442*(1995)364 ; i. barbour , s. hands , j.b .
kogut , m - p .
lombardo , s. morrison , nucl .
phys . * b557*(1999)327 .
lombardo , _ preprint _ hep - lat/9907025 ; s. hands , j.b .
kogut , m.p .
lombardo , s. morrison , nucl .
* b558*(1999)327 . w. kohn , phys .
* 133*(1964)a171 .
wilson , phys . rev . *
k.g . wilson . _ quarks and strings on the lattice _ , in new phenomena in subnuclear physics .
editor a. zichichi . plenum press , new york 1977 .
p. hasenfratz , f. karsch , phys . lett . * b125*(1983)308 .
kogut , h. matsuoka , m. stone , h.w .
wyld , s. shenker , j. shigemitsum d.k .
sinclair , nucl . phys . * b225*[fs9](1983)93 .
i. bender , h.j .
rothe , i.o .
stomatescu , w. wetzel , z. phys . *
i. montvay , g. mnster , _ quantum fields on a lattice_. cambridge univ . press .
gavai , phys . rev . * d32*(1985)519 .
n.w . ashcroft and n.d .
mermin , _ solid state physics _ , saunders college publishing , 1976 .
g. martinelli , g. parisi , r. petronzio , f. rapuano , phys .
* 116b * ( 1982 ) 434 ; c. bernard , t. draper , k. olynyk , m. rushton , phys .
* 49 * ( 1982 ) 1076 ; j. ambjorn , v.k .
mitryushkin , v. g. bornyakov , a. m. zadorozhnyi , phys . lett .
* b225 * ( 1989 ) 153 ; m. lscher , p. weisz , r. sommer , u. wolff , nucl .
* b389 * ( 1993 ) 247 . | we study the linear response to an external electric field of a system of fermions in a lattice at zero temperature .
this allows to measure numerically the euclidean conductivity which turns out to be compatible with an analytical calculation for free fermions .
the numerical method is generalizable to systems with dynamical interactions where no analytical approach is possible .
pacs numbers : 11.15.ha 71.10.fd , 71.20.-b , |
ferromagnetism and singlet superconductivity are two antagonistic long range orderings .
it was demostrated that , however , they can coexist in superconductor / ferromagnet ( sf ) heterostructures . when and how this coexistence takes place , it has been extensively theoretically described in ref.s . in sf
heterostructures a superconductor is in contact to a ferromagnet , and cooper pairs can penetrate into the f layer , inducing an exponentially decreasing superconducting order parameter in the f region .
the length scale is dictated by the ferromagnetic coeherence length @xmath2 , due to the finite lifetime of the cooper pairs in the exchange field of the ferromagnet .
in addition , the diffusion of the ferromagnetic ordering into the s layer(s ) results in a decrease of the superconducting transition temperature @xmath3 ( inverse proximity effect ) of the whole structure . moreover , the interaction between the exchange field and the cooper pairs produces an oscillatory behaviour of the superconducting order parameter in the f region with a spatial periodicity , again on the scale of @xmath2 .
the damped oscillatory behaviour of the superconducting order parameter in the ferromagnetic region is responsible for a nonmonotonic dependence of @xmath3 on the thickness of the ferromagnetic layer . in the specific , and much studied@xcite case of sfs trilayers ,
the competition between s and f orderings gives rise to two possible superconducting ground states , so
called `` @xmath4 '' and `` @xmath5 '' states . in the @xmath4 state
the superconducting order parameter is depressed in the f layer , but no phase difference exists between the s layers . in the @xmath5 state
the superconducting wavefunction shows a @xmath5 phase difference between the s layers , so that the order parameter crosses zero in the f layer .
whether the ground state in a sfs structure is the @xmath4 or @xmath5 , it depends on @xmath2 ( i.e. , on the f material ) and on the ferromagnetic layer thickness @xmath6.@xcite for the very nature of the transition , the main experimental evidence of the @xmath1 transition has been gained through josephson
like experiments.@xcite it is now widely accepted that , by varying the f thickness or the f material , one can observe the @xmath1 transition .
a second , specific aspect of the @xmath1 transition concerns its true thermodynamic nature . since the signature of the transition
is observed or predicted mostly in phase interference effects , one might raise some questions about the true nature of the transition .
recently , it has been experimentally observed@xcite that the transition has a true thermodynamic nature : surface impedance measurements yielded the temperature dependence of the superfluid density in sfs samples with different @xmath6 . in the sample with critical @xmath6 thickness ,
a reentrant jump of the superfluid density was observed , in agreement with theoretical calculations.the experimental behavior was in agreement with theoretical calculations,@xcite and the extended theoretical treatment@xcite confirmed the existence of a jump in superfluid density at the @xmath1 , temperature induced transition.however , while all qualitative features of the theoretical calculations were observed in the measurements of the superfluid density , a few minor quantitative discrepancies existed . as an example
, the temperature induced transition occurred at a ratio @xmath7 different ( by a factor @xmath82 ) than predicted . at a previous stage , it was not possible to ascertain which aspects of the theory ( such as , e.g. , the perfect transparency of s f barriers ) were essential to the phenomenon . thus , continuing the study of the sfs heterostructures is relevant to foster the understanding of the fundamental physics that makes the @xmath1 transition observable .
this paper presents a study of the electrodynamics , of the local structure , and of the compositional profile , of sfs , nb / pdni / nb , heterostructures , successive to a 2 years ageing , in order to give information on the @xmath1 transition after possible deterioration of the samples .
the ageing of the samples helps to stress the effect of structural properties , such as the sharpness of the interfaces ordefects in the composition , which may be crucial to quantitatively match the theory and the experimental evidence.the paper is organized as follows : in sec.[samples ] we summarize the preparation of the sfs heterostructures ; in sec.[mw ] we present microwave electrodynamics in aged samples , as compared to fresh samples ; in sec.[xas ] we present the results of extended x ray absorption fine structure ( xafs ) measurements carried out at the nb , pd and ni @xmath9edges , yielding detailed information on the local structure around each species in our samples ; in sec.[tof ] we explore the compositional depth profile of our samples , with 1 nm resolution , obtaining clear information on the interfaces and on possible interdiffusions ; finally , in sec [ disc ] we discuss the results , and give short conclusions .
nb / pd@xmath10ni@xmath11/nb sfs trilayers have been grown at room temperature on al@xmath12o@xmath13 ( 1120 ) substrates by ultra - high - vacuum dc diode magnetron sputtering .
the base pressure in the chamber was in low 10@xmath14 mbar , while the argon pressure during the deposition was @xmath15=3 @xmath1610@xmath17 mbar .
the sfs trilayers consisted of two nb layers of nominal thickness @xmath18 nm , sputtered at a rate of @xmath19 nm / s , and an intermediate pd@xmath10ni@xmath11 one of different thickness @xmath20 210 nm , deposited at a rate of @xmath21 nm / s .
the deposition rates were calibrated by independent x - ray reflectivity measurements on deliberately deposited single layers of each material sputtered for a given deposition time .
samples are labeled by the letter @xmath22 followed by a number indicating the thickness of the pd@xmath10ni@xmath11 in nm ( e.g. , `` sample d8 '' is the trilayer with @xmath20 8 nm ) . a pure nb film 30 nm thick was also grown for comparison .
more details concerning the sample preparation are described in ref . .
pd@xmath23ni@xmath24 is a weak ferromagnetic alloy widely employed in the framework of the research of the coexistence of s and f orderings , due to the tunability of its exchange energy @xmath25 by a small amount of the ni content . in particular ,
the magnetic and electrical properties of pd@xmath10ni@xmath11 ( hereafter pdni ) were widely investigated in ref . , resulting in an exchange energy @xmath26 mev , a ferromagnetic coherence length @xmath27 nm and a curie temperature @xmath28 k. moreover , it is also worth noticing that a number of works present in the literature suggest that nb / pdni - based systems show relatively high transparency of the s / f interface.@xcite
transition:@xcite @xmath29 ( blue full dots ) , and @xmath30 ( red full dots ) .
inset : sketch of the dielectric resonator and of the film position .
the reentrant behaviour is visible already in the raw data for @xmath31 ( b ) and @xmath32 ( c ) .
reference temperature is @xmath333.19 k. ] microwave measurements are a useful tool to investigate the fundamental properties of superconducting materials , since they give information on both quasiparticles and superfluid density .
we measured the microwave response by means of a cylindrical dielectric resonator , operating at @xmath88 ghz .
a sketch of the experimental setup is reported in the inset of fig.[figexp]a .
the details of the setup and of the procedure have been given elsewhere.@xcite ; ( b ) @xmath34 .
the reference temperature is @xmath33 4.50 k. the flattening of the data indicates that the resonator reached the sensitivity limit , and it is not a direct measure of the normal state . ] in short , the sample was placed at the bottom of the resonator .
measurements of the quality factor @xmath32 and of the resonant frequency @xmath31 yielded the variation of the effective surface resistance @xmath34 and of the effective penetration depth @xmath35 with @xmath36 , according to : @xmath37 and @xmath38 where @xmath39 is a reference temperature and @xmath40 is a geometrical factor that can be computed by standard simulation software .
the most direct information is given by eq .
: in fact , in thin structure of overall thickness @xmath22 as ours , and not too close to @xmath3 , one has:@xcite @xmath41 so that@xcite @xmath42 , with @xmath43 the superfluid density . thus , eq . directly links the measured parameter @xmath31 to the thermodynamic property @xmath43 .
( @xmath34 ) .
data sets have been offset for clarity .
( a ) sample d2 ; @xmath33 3.19 k. the arrows mark the peak of @xmath44 , corresponding to the reentrant jump of the superfluid density at the @xmath1 transition , before ( downward arrows ) and after ( upward arrows ) ageing
. @xmath34 behaves consistently .
( b ) sample d8 , same symbols as in ( a ) ; the arrow marks the appearance of he signature of the @xmath1 transition after ageing ; note also the upward shift of @xmath3 after ageing . ] in a previous work@xcite we have measured @xmath29 in samples d2 , d8 , d9 , and in the nb sample as a reference .
the presence of a sharp _ increase _ of @xmath35 ( that is , a _ decrease _ in @xmath45 ) with decreasing temperature , at @xmath46 , was the evidence for the temperature induced @xmath1 transition , from the high@xmath36 , @xmath4 state , to the low@xmath36 , @xmath5 state .
notably , it was possible to observe the @xmath36induced transition only in the sample with the specific f thickness @xmath20 2 nm .
interestingly , above this thickness a departure from conventional behaviour was observed in the mixed state properties , such as the flux
flow resistivity,@xcite the vortex pinning constant,@xcite and the flux creep factor.@xcite some of these observations were related to a change in the superconducting state .
thus , there is evidence that the @xmath1 transition marks a boundary between different superconducting states , with the @xmath5 state firmly established in samples with large enough @xmath6 ( that is , with the superconducting transition directly into the @xmath5 state ) , and a double transition ( in the superconducting @xmath4 state at @xmath3 , and then in the @xmath5 state at @xmath46 ) in the sample with critical f thickness . in order to show the experimental signature of the @xmath36induced @xmath1 transition , data for @xmath29 and for @xmath47 , as well as the corresponding raw data for @xmath31 and @xmath32
are reported in fig.s [ figexp]b and [ figexp]c .
the data refer to `` fresh '' samples , measured within a few weeks after preparation .
we emphasize that the flattening of the data above @xmath3 is due to the sensitivity limit of the resonator : the flat value should not be taken as the `` normal state '' value . since we have identified the nonmonotonous behavior of @xmath29 as the signature of the @xmath1 transition , it is important to rule out possible spurious effects that can mimic this behavior
. an extensive discussion has been reported previously,@xcite we shortly recall it here .
first , several experimental runs were performed , including runs after entirely disassembling and reassembling the microwave resonator : the data completely overlapped .
second , the same jump in @xmath29 obtained by direct measurements in zero magnetic field ( as those here reported ) , was obtained by estimating the effective penetration depth from the condensation energy derived from magnetic field dependent microwave measurements ( a completely different experiment ) : the results from the different experiments were in excellent agreement . moreover ,
multiple transitions are ruled out by the fact that they would lead to a stepwise temperature dependence of @xmath29 , and not to a reentrance ( note that the data for @xmath47 point to the same conclusion ) .
a more subtle spurious effect may reside in the stray field of the magnetic layer itself and/or of magnetic small particles embedded in the superconducting matrix .
such phenomena have been widely studied,@xcite with the largest influence on the magnetoresistance and the magnetization in external magnetic fields .
we focus here on the few effects that can be observable in our zero field measurements . in superconducting films with different local concentration of ferromagnetic nanoparticles
, a multi state resistance value could be observed.@xcite the effect was enhanced by an external magnetic field ( which is not our case ) , and most importantly it would similarly give , in our sample averaged measurements , a multi step transition both in @xmath47 as well as in @xmath29 . the effect would be similar to the transition of a multiphase superconductor , while the observed behavior is a reentrance of the superconducting properties .
a second possible effect comes from the stray field due to the coupling of the external ferromagnetic layers in fsf heterostructures:@xcite the stray field could locally exceed the critical field . in our sfs heterostructures , that contain a single f layer ,
this phenomenon is likely to be much weaker .
moreover , the effect on our sample averaged measurements would again be similar to a multiphase superconductor , thus giving rise to a multi step or a broadened transition .
thus , we believe that spurious effects are unlikely to produce a reentrance in the superconducting propeorties , in our experimental configuration and with the check we have made . as it will be largely discussed in the following , we anticipate that repetition of the measurements on the same samples after 24 months yielded strongly different results ( while short term repetition of the measurements gave identical results , as discussed in ref . ) .
since aim of this paper is to understand this ageing effect and gather informations from joint electrodynamics , structural and chemical investigation , we performed a preliminary repetition of the measurements on the reference nb sample , as grown and after ageing .
the results are presented in fig .
[ fignb ] , in terms of @xmath29 , fig.[fignb]a , and @xmath47 , fig.[fignb]b . as it can be seen , after a rather long time of 24 months the measurements are practically identical .
thus , we conclude that any ageing effect that should be observed must not be ascribed to any deterioration of the nb itself , but only to alterations of the pdni layer or of the f / s interfaces fig .
[ figmwaged ] reports the main experimental findings of this section .
there , we report and compare @xmath29 and @xmath48 obtained for the fresh and aged trilayers d2 and d8 .
several effects emerge .
first , @xmath3 of the sample d2 is little changed ( a small shift upward , @xmath49 50 mk , might be inferred ) , while in sample d8 @xmath3 clearly shifts _ upward _ with ageing : this result is in clear contradiction with any nave interpretation of the ageing effects in terms of just a deterioration of the samples .
the involved phenomena are clearly more complex .
the second important effect is the shift at higher @xmath36 of the @xmath1 transition in sample d2 , where it is accompanied by a broadening and weakening , as indicated by the very faint remaining features , and the emergence of the fingerprint of the transition in sample d8 , although in a broader fashion than in the fresh sample d2 .
it is then clear that the ageing produce changes in the pdni layer , that affect the magnetic properties and then the overall superconducting order parameter , in a nontrivial fashion .
thus , before any discussion of the electrodynamics data a structural and compositional analysis is in order .
the following sections are devoted to this task .
x - ray absorption spectroscopy ( xas ) is a chemical selective probe@xcite that we used to investigate the local structure around nb , ni and pd atoms in the aged samples and to individuate the modifications of the local structure and of the composition with respect to the nominal parameters . to this aim
we have collected nb @xmath9edge xafs spectra at the esrf bm08 beamline@xcite ( grenoble , france ) , and pd and ni @xmath9edge xafs spectra at the elettra xafs@xcite beamline ( trieste , italy ) .
xas spectra were measured in fluorescence geometry , keeping the samples at the liquid nitrogen temperature in order to reduce the thermal contribution to structural disorder .
nb @xmath9edge xas spectra were collected using an ultrapure ge multidetector ( 13 elements ) , pd and ni @xmath9edge xas spectra were measured using a ketek silicon drift detector . + ( a ) [ cols="^,^,^ " , ] in fig .
[ fignbspectra ] we report the experimental @xmath50 in all samples measured ( red dots ) , together with the fits ( continuous lines ) .
the good fittings demonstrate that the nb bcc bulk structure is a good model for the average structure around nb in the s layers .
the best fit parameters are in agreement with nb bulk structure , in particular @xmath51 3.3 is unchanged within the uncertainty for all the investigated samples , demonstrating that the bcc bulk - like phase is preserved for the nb layers in all the samples measured . weighted xafs data at the ni edge ( coloured dots ) and best fits ( full black lines ) .
parameters are reported in the table .
it is found that ni does not bond to pd only , but it is also found close to nb , in the ninb@xmath52 alloy . + * * in pdni , @xmath53 is not a free parameter ( see text ) ; + * * * in the ni rich phase , @xmath54 is kept constant ( see text).,title="fig : " ] + lllll + & + & @xmath53 & @xmath55 & @xmath56 & @xmath57 + & ( ) & ( @xmath58 ) & ( @xmath58 ) & ( % ) + d2 & 3.80 * & 8(3 ) & 19(3 ) & 65(5 ) + d4 & 3.84 * & 11(3 ) & 14(3 ) & 80(6 ) + d8 & 3.83 * & 6.2(3 ) & 12(3 ) & 76(4 ) + d10 & 3.82 * & 5.3(3 ) & 16(3 ) & 72(4 ) + & + & @xmath59 & @xmath55 & - & @xmath60 + & ( ) & ( @xmath58 ) & & ( % ) + d2 & 3.50(2 ) & 2.0 * * & & 19(3 ) + d4 & 3.51(2 ) & 2.0 * * & & 8(2 ) + d8 & 3.50(2 ) & 2.0 * * & & 8(2 ) + d10 & 3.50(2 ) & 2.0 * * & & 9(2 ) + weighted xafs data at the pd edge ( coloured dots ) and best fits ( full black lines ) .
parameters are reported in the table .
it is found that pd does not bond to ni only , but it is also found close to nb , in the nb@xmath13pd alloy . + * * in pdni , @xmath53 is not a free parameter ( see text ) .
, title="fig : " ] + lllll + & + & @xmath53 & @xmath55 & @xmath56 & @xmath57 + & ( ) & ( @xmath58 ) & ( @xmath58 ) & ( % ) + d2 & 3.83 * & 6.4(5 ) & 15(6 ) & 76(4 ) + d4 & 3.82 * & 5.2(4 ) & 12(6 ) & 72(5 ) + d8 & 3.84 * & 4.4(4 ) & 12(5 ) & 80(3 ) + d10 & 3.83 * & 4.2(3 ) & 4.1(3 ) & 76(3 ) + & + & @xmath61 ( ) & @xmath55 & @xmath56 & @xmath60 + & ( ) & ( @xmath58 ) & ( @xmath58 ) & ( % ) + d2 & 3.19(2 ) & 11 * * & 16 * * & 65(5 ) + d4 & 3.17(1 ) & 11(4 ) & 16(5 ) & 60(5 ) + d8 & 3.16(1 ) & 12(4 ) & 15(4 ) & 50(3 ) + d10 & 3.16(1 ) & 9(3 ) & 13(4 ) & 51(3 ) + the exafs signal in multilayers appears attenuated with respect to pure nb film ( fig.[fignbspectra ] ) .
this finding must be attributed to a somewhat larger disorder of the nb structure in the trilayers , suggesting some influence on the s layer by the f layers .
the quantitative analysis ( best fit parameters ) shows that the nb
nb interatomic distances are unchanged in the trilayers .
the structural disorder ( @xmath62 parameters ) in the trilayers is higher than in pure nb film , but it is similar in all aged trilayers ( within the uncertainty ) , suggesting that the disorder induced in the nb layers is weakly dependent on the pdni film thickness .
further understanding about the mechanisms giving rise to the structural disorder and to the modification of the superconducting response of the heterostructures can be obtained looking at the local atomic structure around pd and ni atoms .
ni and pd @xmath9edge xas spectra were measured on 4 samples with increasing pdni layer thickness @xmath6 , from 2 to 10 nm . to gain more insight on the trend of the disorder around pd and ni , here
we have extended the investigation to the sample with @xmath20 10 nm .
pd and ni have the fcc structure ( fig.[fig : strutt]b ) with lattice parameters @xmath63=3.89 and @xmath59=3.52,@xcite respectively .
literature data reports that the pd@xmath24ni@xmath23 alloys have the fcc structure with lattice parameters ranging from @xmath64=3.76 ,@xcite to @xmath65=3.87 .@xcite the analysis of pd and ni @xmath9edge spectra is relatively complex and detailed with care in the appendix a , here we focus on the structural findings .
the @xmath66weighted pd and ni @xmath9edge experimental exafs spectra @xmath67 , along with the best fit curves , are shown in fig.s [ figniexafs ] and [ figpdexafs ] for all the four trilayers , and the resulting parameters for the best fits are reported in the tables accompanying fig.s [ figniexafs ] and [ figpdexafs ] .
pd xafs spectra present higher statistical noise for thinner samples , in particular d2 and d4 samples had to be analysied in a restricted @xmath66range .
the amplitude of the exafs signal steadily increases with increasing @xmath6 .
this is especially evident on the ni @xmath9edge data .
it is then a qualitative observation that in aged samples the disorder around pd and ni is larger in thinner pdni layers .
this is a suggestion that the degradation affects mainly the regions close to the interface since should it be a homogeneous volume effect no trend would be observed with increasing the pdni thickness .
the analysis of ni xafs data is consistent with the pd@xmath68ni@xmath23 fcc model , with a small fraction of ni ( about 10 % ) in a ni rich phase .
the best fit parameters suggest a pd@xmath68ni@xmath23 phase with @xmath57 between 0.7 and 0.8 for all thicker samples ( slightly different from the nominal @xmath69 0.84 ) , while it decreases around 0.65 for the thinner sample in which also a large fraction of ni rich phase is found .
the disorder factors roughly increases for thinner samples .
the results of the pd @xmath9edge analysis are consistent with ni data : in particular the pd@xmath68ni@xmath23 phase appears richer in ni for all the samples with respect to the nominal composition and the structural disorder increases in thinner samples .
moreover all the samples depict a fraction of pd in the nb rich phase , larger in sample d2 .
we must notice that the contributions from the ni and nb neighbours are often in anti
phase , this could give rise to some systematic errors on the final values .
moreover , the imposed constraints may determine a systematic inaccuracy .
thus , while absolute values for @xmath57 may well be affected by systematic inaccuracy , we are confident about the relative trend resulting from the exafs analysis as a function of @xmath6
. in particular , our findings suggests that structural and compositional disorder occurs within the f layers giving rise to a deviation of the local atomic structure from the ideal one .
the origin of these effects resides in the f / s interfaces , as they are larger for thinner samples .
the results of ni and pd exafs data analysis consistently suggest that the pd@xmath68ni@xmath23 phase is systematically richer in ni with respect to the x=0.84 nominal composition .
a relatively large fraction of pd occurs in a nb rich phase , suggesting pd atoms migrate into the s layers .
this finding is consistent with our previous tem analysis,@xcite which showed the presence of heavy ions ( arguably , pd ) within nb , and in particular in the bottom nb layer .
a fraction of ni atoms goes in a ni rich phase , suggesting that migration of pd into nb not only would leave ni richer pd@xmath68ni@xmath23 , but can also give rise to almost pure ni regions in the sample .
from the exafs study it emerges that substantial modifications occurred in the pdni , f layers of aged samples , with possible effects on the nb , s layers .
we stress that the magnetic properties of the pdni alloy change with the composition,@xcite so that any information on the actual composition of the magnetic layer is very relevant to the understanding of the electrodynamics data in order to clarify further the changes , we performed a detailed analysis of the compositional profile in trilayers , using the tof
sims technique as described in the next section .
tof sims ( time of flight secondary ion mass spectrometry ) analysis is a very powerful technique to assess the composition of a given material@xcite by controlled sputtering .
the capability to detect tiny traces of chemical species within small areas ( less than 0.01 mm@xmath58 ) , and with a depth resolution that approaches the nm , makes it a particularly suitable tool for the investigation of nm
thick layers , such as our sfs heterostructures , in the search for anomalies at the interfaces between layers,@xcite or for contaminants between and inside the layers.@xcite the measurements were performed using a tof.sims v ( ion - tof gmbh , mnster , germany ) apparatus .
sims depth profiling have been performed in the dual - beam mode .
we have analyzed samples d2 , d4 , d8 , in order to directly compare the results to the exafs data and to detect systematic effects ( if any ) with @xmath6 .
the sputtering was performed using a 0.5 kev cs@xmath70 ion beam rastered over an area of 300@xmath16300 @xmath71m@xmath58 .
a 30 kev primary bi@xmath70 ion source rastered over a scan area of 100@xmath16100 @xmath71m@xmath58 constitutes the analysis beam . both ion beams were impinging the sample surface forming an angle of 45 with the surface normal . due to the relative complexity of the heterostructure , which contains nb , pd , ni , al ( as part of the substrate , sapphire ) , and almost certainly nb oxyde , we analyzed al@xmath70 , ni@xmath70 , nb@xmath70 , nbo@xmath70 and @xmath72pd@xmath70 signals as a function of the depth in our samples .
all the measurements were repeated in three different locations for each sample , to check the uniformity of the results .
the sputtering time was converted to depth using a measurement of the overall thickness with a profilometer , the appearance of nb@xmath70 as zero depth and the sharp increase of al@xmath70 as a measure of the substrate level . in connection to the detection of al@xmath70 , we note that , since we were also interested in evaluating the thickness of each layer , we use very little or no presputtering . thus , close to the surface we detect all the surface contaminants , and al among them : clearly , this is not due to the substrate . before the presentation of the results , we remind that tof
sims data relative to different species can not be quantitatively compared : different intensities of the signal in different species are not necessarily a measure of the relative abundance .
moreover , primary ions , during their impact on the target surface , give rise to implantation phenomena which produce a persistency of the detection of a species even beyond its existence in the material .
thus , one should give particular care to the increase and decrease of the traces .
we first assess the uniformity of the composition of the layers in different locations of the surface .
we found very consistent results in all samples .
as an illustration , in fig .
[ fig : unifo ] we report the depth profiles for nb@xmath70 , ni@xmath70 and @xmath72pd@xmath70 in sample d8 .
it is seen that all traces show the same dependence on the depth : this is a demonstration that the trilayer is homogeneous over planes parallel to the surface .
moreover , it should be especially noted that ni@xmath70 and @xmath72pd@xmath70 traces track the same behaviour . from observation of fig.[fig : unifo ] , we note that the effective thickness of the layers is consistently smaller than the nominal value .
let us focus momentarily on the nb@xmath70 traces : using the kink of the decrease as a marker for the boundary of the first ( outmost ) nb layer with pdni , we estimate @xmath73 11 nm for the first layer . using the increase of the same trace at a depth @xmath74 16 nm as the marker for the second interface
, we have again @xmath73 11 nm for the second ( inner ) layer , to be compared with nominal value of 15 nm .
we will see in the next figures that , as it is very reasonable , this is due to oxydized nbo layers ( with o coming from the air for the outer layer , and from the substrate for the inner layer ) .
we note that also the pdni layer is thinner than the nominal value . using the same criteria to evaluate the thickness
, we estimate @xmath75 6 nm , as compared to the nominal value of 8 nm . the smooth , slow decrease of the traces above the depth of 16 nm , together with the exafs data ,
indicate that the smaller thickness of the f layer is likely to be due to diffusion processes at the lower interface pdni / nb .
the general features here uncovered are found in all samples . in order to discuss more thoroughly the composition of the aged structures
, we now report for each sample the full set of traces ( taken in a single location to avoid crowding ) .
figure [ fig : tofsims ] reports in panels ( a ) , ( b ) and ( c ) the traces for al@xmath70 , ni@xmath70 , nb@xmath70 , nbo@xmath70 and @xmath72pd@xmath70 as a function of the depth in samples d2 , d4 and d8 , respectively . the ni@xmath70 and @xmath72pd@xmath70 signals are close to the sensitivity level , but nevertheless they show the existence of the intermediate pdni layer .
the comparison with the nb@xmath70 trace allows to evaluate the pdni thickness as @xmath76 2 , 3.5 , 6 nm in samples d2 , d4 , d8 , respectively . ,
ni@xmath70 , nb@xmath70 , nbo@xmath70 and @xmath72pd@xmath70 signals as a function of the depth in samples d2 , d4 , d8 ( from top to bottom ) . in all cases
it is seen that the nb layers have an actual thickness of @xmath8 10 nm . ] the nb@xmath70 signals decrease , but remain appreciable in the region corresponding to the pdni layer .
this is likely to be due to the collision processes , responsible for the re implantation of ions in the sample .
however , we can not exclude that part of the signal comes from a blurred interface between nb and pdni .
we note that the `` tails '' of the ni@xmath70 and @xmath72pd@xmath70 signals are particularly long in sample d4 . taking into account that the measurements for pd and ni are very close to the sensitivity limit
, we believe that this result is consistent with exafs data indicating diffusion of pd into the nb ( inner ) layer .
nbo is the only significant contaminant in our samples : it is mainly present in a region which extends for @xmath77 5 nm in depth , both at the air / nb and nb / sapphire interfaces , whence the obvious origin of oxygen .
however , we note that there is a second , smaller peak of the nbo@xmath70 trace at the pdni / nb ( inner layer ) interface , most likely originating in the deposition process . summarizing , from tof
sims measurements in aged heterostructures we have obtained that ( i ) the trilayers are very homogeneous over planes parallel to the surface , ( ii ) the actual nb ( s ) thickness is @xmath78 ( 11 @xmath79 1 ) nm , ( iii ) the actual pdni ( f ) thickness is close , but slightly reduced with respect to nominal values , ( iv ) the interfaces , and in particular the f / s ( inner layer ) interface , is somewhat blurred over 1 - 2 nm . in the next and last section we discuss all the information gathered from all the measurements up to now presented .
the whole body of data collected allows for several remarks on the effect of ageing .
the following discussion takes into account the effects observed on the superconducting and the @xmath1 transition after ageing , that is the shift to _ higher _ temperature of @xmath3 in samples d2 and d8 , the shift of @xmath80 at higher @xmath36 and the broadening of the transition in sample d2 , and the appearance of the signature of the @xmath1 transition in sample d8 .
we first argue that the effects of the nb oxidisation do not play any role .
in fact , the electrodynamics measurements in nb do not show any effect at all with ageing .
we note that even a simple reduction of the thickness of the superconducting layer should bring detectable changes : since the effective surface resistance in thin samples is@xcite @xmath81 , a reduction of the thickness @xmath22 would immediately bring a scale factor in the measured @xmath82 , which is not seen in the data of fig .
[ fignb ] .
thus , we conclude that the presence of nbo stems from oxidisation during deposition , for the interface nb / sapphire , and upon exposition to the atmosphere , for the outermost layer , and that it is not the result of ageing . as a consequence ,
the actual nb thicknesses of our samples have to be taken as @xmath73 20 nm in the pure nb film , and @xmath78 10 nm in the trilayers .
the changes in the pdni layer are more relevant to the @xmath1 transition .
let us focus first on sample d2 .
according to the previous findings,@xcite the heterostructure entered the superconducting state at @xmath3 in the @xmath4 state , and only at lower @xmath80 the @xmath1 transition occurred , to a _
weaker _ superconducting state ( that is , with a smaller superfluid density ) . by changing the effect of the f layer ( increased disorder , less clean due to mixing with the s material ) ,
the @xmath1 transition is broadened .
the tof sims data taken in different locations of the sample stand against an explanation of the broadening in terms of local inhomogeneity .
instead , there is a clear inhomogeneity along the depth of the structure .
thus , the weakening and broadening of the @xmath1 transition should be ascribed first to the blurring of the interfaces ( most notably , the f / s interface with the inner s layer ) .
exafs also points to a change in the alloy . with a larger amount of ni
, a shorter @xmath2 is expected .
this effect can counterbalance the deterioration of the sharpness of the interfaces and the increase of disorder , leaving a faint trace of the @xmath1 transition .
we note that theoretical calculations in sfs heterostructures with a modulation of the interface predict a weakening of the sharpness of the @xmath1 transition , in terms of a change of the transition from first to second order.@xcite in sample d8 the increased disorder has a smaller effect , due to the larger f thickness . in this case , we believe that the strongest alteration is the change of the effective f thickness , from 8 nm to less than 6 nm . this effect can not be easily mapped onto a clean system , but the appearance of the @xmath1 fingerprint indicates that the sample has sufficiently decreased its effective f thickness to enter the region where the superconducting state at @xmath3 is the @xmath4 state .
the fact that @xmath3 increased significantly ( by approximately 0.3 k , with @xmath83 3.8 k in the fresh sample ) also points to a smaller effect of the f layer . in summary
, the reentrance of the superconducting order parameter , fingerprint of the transition 0-@xmath5 , was investigated by means of microwave measurements in `` fresh '' and `` aged '' sfs trilayers as a function of the f layer thickness .
the structural and compositional properties of the aged samples have been studied by means of exafs and tof sims analysis . increased local disorder , and formation of nbpd and ninb alloys ,
were found in aged samples .
sims indicated that the interfaces were most affected . despite severe effects in the ordering of the species in the f layers and in the quality of the sf interfaces , the temperature induced @xmath1 transition is still detected .
thus , the assumption of ideal interfaces does not seem to be an essential aspect of the theoretical description , and other physical properties play a stronger role : e.g. , the exchange field and/or the ferromagnetic coherence length.@xcitemoreover , our data suggest that an `` effective '' f thickness can include non negligible disorder and nonideal interfaces , at least as the most relevant correction .
lt thanks `` fondazione roma '' for financial support to the surface analysis laboratory at roma tre .
the analysis of pd and ni @xmath9edge spectra required some care because the data statistics is generally lower compared with nb data , partially because of the smaller amount of absorbers that reduces the fluorescence signal . moreover , higher structural disorder is expected due to the confinement in a thinner layer and to interface effects .
in addition , the pd and ni atoms are randomly distributed on the fcc lattice : this raises the number of contributions to be considered , as each shell must include scattering from both ni and pd atoms .
a preliminary analysis suggested a local structure around pd and ni different from pure pd@xmath84ni@xmath85 alloy , and pointed to some interdiffusion effect ( nb rich pdnb phase ) and ion segregation ( ni rich phase ) that must be considered carefully to achieve a reliable quantitative interpretation of the structural and compositional results . due to the number of contributions to be considered in the analysis it is required to impose physical constraints on the fitting parameters , in order to reduce the correlation effects and to avoid unphysical results .
we apply the following model : + * i*. we restrict the analysis to the first two coordination shells of the pdni phase ( up to about 3.9 ) , each shell including both ni and pd neighbours ; + * ii*. the total coordination numbers of each shell were constrained to the fcc structure ( n@xmath86=12 , n@xmath87=6 ) .
a free parameter @xmath57 is defined , providing the number of pd ( @xmath88 ) and ni ( @xmath89 ) neighbours for the @xmath90th shell ; + * iii*. @xmath91 , @xmath92 were further multiplied by a factor @xmath93 , @xmath60 being a free parameter in the fitting representing the fraction of absorber in a phase different from pd@xmath24ni@xmath23 ( see below ) ; + * iv*. we assume that the lattice parameter @xmath94 of the pd@xmath24ni@xmath23 alloys has a linear behaviour as a function of the composition @xmath57 for @xmath95 .
the parameters are fixed by linear regression trough @xmath64 , @xmath65 and @xmath63 values .
notice that @xmath96 is not a free parameter but constrained to @xmath57 ; + * v*. for each shell the disorder factors , @xmath97 , were kept equal for pd and ni neighbour .
preliminary tests suggest that additional contributions must be considered in the analysis of pd and ni xafs , other than pd@xmath24ni@xmath23 phase . in this respect ,
pd and ni data behave differently : for pd @xmath9edge data , an additional contribution is consistent with a nb rich phase .
we assume this phase having bcc structure accordingly to literature data for nb
rich nb@xmath68pd@xmath23 phases.@xcite we included the first two bcc coordination shells in the analysis , the coordination numbers were weighted by the free parameter @xmath60 , giving the fraction of pd into the nb rich phase .
for the ni @xmath9edge data an additional contribution is consistent with ni
ni coordination shell around 2.5 , consistent with fcc ni phase ( @xmath59=3.52 ) .
the fraction of ni in the ni rich phase is refined using a free parameter @xmath60 .
the disorder factor @xmath98 was kept fixed to avoid the strong correlation with the parameter @xmath60 .
a. i. buzdin , _ rev .
_ * 77 * , 935 ( 2005 ) .
a. a. golubov , m. yu .
kupriyanov , and e. ilichev , _ rev .
phys . _ * 76 * , 411 ( 2004 ) . v. v. ryazanov , v. a. oboznov , a. yu .
rusanov , a. v. veretennikov , a. a. golubov , and j. aarts , phys .
lett . * 86 * , 2427 ( 2001 ) .
t. kontos , m. aprili , j. lesueur , f. gent , b. stephanidis , and r. boursier , phys .
* 89 * , 137007 ( 2002 ) .
h. sellier , c. baraduc , f. lefloch , and r. calemczuk , _ phys .
b _ bf 68 , 054531 ( 2003 ) .
v. a. oboznov , v. v. bolginov , a. k. feofanov , v. v. ryazanov , and a. i. buzdin , phys .
lett . * 96 * , 197003 ( 2006 ) .
n. pompeo , k. torokhtii , c. cirillo , a. v. samokhvalov , e. a. ilyina , c. attanasio , a. i. buzdin , and e. silva , _ phys .
b _ * 90 * , 064510 ( 2014 ) . a. v. samokhvalov and a. i. buzdin , _ phys . rev .
b _ * 92 * , 054511 ( 2015 ) c. cirillo , e. a. ilyina , c. attanasio , supercond .
* 24 * , 024017 ( 2011 ) . c. cirillo , s. l. prischepa , m. salvato , c. attanasio , m. hesselberth , and j. aarts , _ phys .
b _ * * 72 * * , 144511 ( 2005 ) .
n. pompeo , k. torokhtii , e. silva , _ meas .
* 14 * , 164 ( 2014 ) . k. torokhtii , c. attanasio , c. cirillo , e. a. ilyina , n. pompeo , s. sarti , and e. silva , physica c * 479 * , 140 ( 2012 ) .
e. silva , m. lanucara , and r. marcon , supercond .
* 9 * , 934 ( 1996 ) .
n. pompeo , r. marcon , l. mchin , and e. silva , supercond .
sci . technol . * 18 * , 531 ( 2005 ) .
m. tinkham , _ introduction to superconductivity _ , dover publications , 2004 .
e. silva , n. pompeo , and s. sarti , supercond .
technol . *
* 24 * * , 024018 ( 2011 ) .
k. torokhtii , n. pompeo , c. meneghini , c. attanasio , c. cirillo , a. ilyina , s. sarti , e. silva , _ j. supercond . nov .
* 26 * , 571 ( 2013 ) .
k. torokhtii , n. pompeo , c. meneghini , c. attanasio , c. cirillo , e. a. ilyina , s. sarti , e. silva , `` transport properties of nb / pdni / nb trilayers at microwave frequencies '' , _ physics and engineering of microwaves , millimeter and submillimeter waves ( msmw ) , 2013 international kharkov symposium on _ , pp.658660 , 2328 june 2013 , doi : 10.1109/msmw.2013.6622165 ( 2013 ) .
n. pompeo , k. torokhtii , e. silva , _ j. supercond
_ * 28 * , 1097 ( 2015 ) .
n. pompeo , k. torokhtii , c. meneghini , s. mobilio , r. loria , c. cirillo , e. a. ilyina , c. attanasio , s. sarti , and e. silva , _ j. supercond . nov . magn . _ * * 26 * * , 1939 ( 2012 ) .
d. stamopoulos , m. pissas , and e. manios , _ phys .
b _ * * 71 * * , 014522 ( 2005 ) .
d. stamopoulos , e. manios , and m. pissas , _ phys .
b _ * * 75 * * , 184504 ( 2007 ) .
g. bunker , introduction to xafs , cambridge university press , berlin , 2010 .
s. pascarelli , f. boscherini , f. dacapito , j. hrdy , c. meneghini and s. mobilio ,
_ j. synchrotron radiat . _
* 3 * , 147 ( 1996 ) .
a.di cicco , g. aquilanti , m. minicucci , e. principi , n. novello , a. cognigni , l. olivi , _ j. phys . : conf .
ser . _ * 190 * , 012043 ( 2009 ) . c. meneghini , s. di matteo , c. monesi , t. neisius , l. paolasini , s. mobilio , c. r. natoli , p. a. metcalf , and j. m. honig , _ phys .
b _ * 72 * , 033111 ( 2005 ) . c. meneghini , f. bardelli , s. mobilio , _ nucl . instrum .
methods b _ * 285 * , 153 ( 2012 ) . c. battocchio , c. meneghini , i. fratoddi , i. venditti , m. v. russo , g. aquilanti , c. maurizio , f. bondino , r. matassa , m. rossi , s. mobilio , g. polzonetti , _ j. phys
c _ , * 116 * , 19571 ( 2012 ) .
ankudinov , b. ravel , j.j .
rehr , and s.d .
conradson , _ phys .
b _ * 58 * , 7565 ( 1998 ) .
wyckoff , r. w. g. _ crystal structures _ , interscience publishers , new york , 1963 , chap . 1 , pp .
783 f. lihl , h. ebel , w. baumgartner , _ zeitschrift fuer metallkunde _ * 62 * , 42 ( 1971 ) .
t. b. flanagan , da wang , j. d. clewley , h. noh , _
j. alloys compd . _
* 305 * , 172 ( 2000 ) .
h. oesterreicher , j. clinton , _ j. solid state chem . _
* 17 * , 443 ( 1976 ) ; b. c. giessen , n. j. grant , d. p. parker , r. c. manuszewski , r. m. waterstrat , _ metallurgical transactions a _ * 11 * , 709 ( 1980 ) . c. meneghini and t. neisius , private communication .
j. aarts , c. attanasio , c. bell , c. cirillo , m. flokstra , j. m. van der knaap , _ superconductor / ferromagnet hybrids : bilayers and spin switching _ , in : nanoscience and engineering : supeconductivity , book series : nanoscience and technology , 323 , ( 2010 ) .
a. benninghoven , _ angew .
int . edit .
_ , * 33 * , 1023 ( 1994 )
. v. sorianello , a. de iacovo , l. colace , a. fabbri , l. tortora , e. buffagni and g. assanto , _ appl .
* 101 * , 081101 ( 2012 ) .
f. zurlo , e. di bartolomeo , a. depifanio , v. felice , i. natali sora , l. tortora , s. licoccia _
j. power sources _ * 271 * , 187 ( 2014 ) .
a. i. buzdin , _ phys .
b _ * 72 * , 100501 ( 2005 ) v. n. kushnir , s. l. prischepa , j. aarts , c. bell , c. cirillo , and c. attanasio , _ eur
j. b _ * 80 * , 445 ( 2011 ) . | we have studied the temperature induced @xmath0 thermodynamic transition in nb / pdni / nb superconductor / ferromagnetic / superconductor ( sfs ) heterostructures by microwave measurements of the superfluid density . we have observed a shift in the transition temperature with the ageing of the heterostructures , suggesting that structural and/or chemical changes took place .
motivated by the electrodynamics findings , we have extensively studied the local structural properties of the samples by means of x - ray absorption spectroscopy ( xas ) technique , and the compositional profile by time of flight secondary ion mass spectrometry ( tof sims ) .
we found that the samples have indeed changed their properties , in particular for what concerns the interfaces and the composition of the ferromagnetic alloy layer .
the structural and compositional data are consistent with the shift of the @xmath1 transition toward the behaviour of heterostructures with different f layers .
an important emerging indication to the physics of sfs is the weak relevance of the ideality of the interfaces : even in aged samples , with less than ideal interfaces , the temperature induced @xmath1 transition is still detectable albeit at a different critical f thickness . |
despite a big experimental effort done in the past years , the origin of the ultra high energy cosmic rays ( uhecrs ) , i.e. with energies larger than @xmath3 ev , is still unknown .
the pierre auger @xcite and telescope array @xcite observatories , are currently taking data in this energy range .
the southern hemisphere pierre auger observatory , in the province of mendoza , argentina , is the largest cosmic ray observatory in the world .
the telescope array observatory , placed in utah , usa , is the largest in the northern hemisphere .
the uhecr energy spectrum presents two main features , the ankle , found at an energy of @xmath4 ev , which consists in a hardening of the flux , and a suppression at @xmath5 ev @xcite .
note that , even though the spectra observed by the pierre auger and telescope array observatories present some differences , these two features are observed in both measurements .
uhecrs generate atmospheric air showers when they interact with the molecules of the atmosphere .
these air showers can be observed by using ground detectors and/or fluorescence telescopes .
the ground detectors detect the secondary particles , produced during the shower development , that reach the ground . whereas , the fluorescence telescopes detect the fluorescence light emitted by the interaction of the secondary charged particles with the molecules of the atmosphere .
the pierre auger and telescope array observatories have both types of detectors , and a subsample of the events are observed by both detector types .
the arrival direction , the energy , and the composition of the primary uhecr have to be inferred from the air shower observations . in particular , the composition is obtained by comparing observables sensitive to the primary mass with simulations of the atmospheric air showers .
this is subject to large systematic uncertainties because the hadronic interactions at the highest energies are unknown .
there are models that extrapolate the lower energy accelerator data to the highest energies .
some of these models have been recently updated with the large hadron collider data . as a consequence
, the differences among the predictions of the different models decreased but did not disappear .
one of the parameters most sensitive to primary mass is the atmospheric depth of the shower maximum , @xmath6 .
this parameter can be reconstructed with the data taken by fluorescence telescopes .
the composition analyses based on @xmath6 observed by auger , done by using the updated hadronic interaction models , show a decrease of the primary mass from @xmath0 ev up to @xmath1 ev and then , beyond that energy , the mass becomes increasingly heavy @xcite . on the other hand ,
the @xmath6 data obtained by telescope array is consistent with protons @xcite .
however , it has been recently shown that the composition data observed by the two experiments are consistent @xcite .
it is worth mentioning that , the statistics of telescope array is smaller than the one corresponding to auger .
the interpretation of the uhecr spectrum depends strongly on the mass composition .
if protons are the dominant component , the ankle is originated by pair production in the interaction of the protons with the low energy photons of the extragalactic background light ( ebl ) and cosmic microwave background ( cmb ) , during propagation through the intergalactic medium .
the suppression observed at the highest energies can originate by the photopion - production of protons interacting with the cmb photons , by the intrinsic cutoff of in the sources , or by a combination of both effects .
the proton model ( known as the dip model ) of the uhecr spectrum have been extensively studied in the literature ( see refs .
@xcite ) .
on the other hand , if the uhecr spectrum is composed by heavy nuclei besides protons , the ankle can be interpreted as the transition between the galactic and extragalactic cosmic rays ( see ref .
@xcite for a review ) . however , this possibility is disfavoured by the auger data @xcite , large scale anisotropy studies show that the extragalactic component should continue below the ankle . recently , a new light extragalactic component , that dominates the flux below the ankle , originated in different sources than the ones responsible of the flux at the highest energies , has been proposed @xcite .
this low energy component could also originate in the photodisintegration of nuclei with the photon fields in the vicinity of the acceleration region or the source @xcite . in these scenarios
the suppression is due to the photodisintegration process undergone by the nuclei when they interact with photons of the ebl and cmb , to an intrinsic spectral cutoff in the injected spectrum , or to a combination of both effects .
besides composition information , the observation of secondary gamma rays and neutrinos , generated in the interactions undergone by the cosmic rays during propagation through the universe , can constrain the different models of the uhecr spectrum . in particular , a smaller number of gamma rays and neutrinos are predicted in scenarios with larger fractions of heavy nuclei at the highest energies . in ref .
@xcite the dip model has been rejected at 95% cl by using the upper limit on the neutrino spectrum obtained by the icecube experiment @xcite . in that work the source evolution
is assumed to be the one corresponding to the star formation rate times @xmath7 where @xmath8 is the redshift and @xmath9 is a parameter fixed by fitting telescope array data .
however , it is also shown that models with no emission at @xmath2 can not be rejected with present neutrino data .
extragalactic gamma - ray background ( egb ) observations impose quite restrictive constrains on models of the uhecr spectrum @xcite .
the egb has recently been measured , from 100 mev to 820 gev , by fermi - lat @xcite .
part of the egb originates in gamma - ray point sources .
the gamma rays generated by the propagation of uhecrs can contribute to a diffuse component . by using the 2fhl catalog @xcite of 360 point sources ( mostly blazars ) detected by fermi - lat
, it has been shown that @xmath10 of the integrated egb spectrum from 50 gev to 2 tev originates in point sources @xcite . by using this result , in ref .
@xcite it is found that only a group of nearby sources , contributing in the energy range below the ankle , can be responsible for the light component observed in that energy range .
also , in ref .
@xcite , a more restrictive upper limit to the energy density of the electromagnetic cascades , that are developed in the intergalactic medium , is found by using this new result . in this work
we study the impact of this new estimation on proton models of uhecrs . instead of using the upper limit to the energy density found in ref .
@xcite , we calculate an upper limit to the integrated gamma - ray flux , in the energy range from 50 gev to 2 tev , that does not originate in point sources .
this upper limit is used to study the constraints that this new result imposes on proton models of the uhecrs spectrum , assuming a more relaxed source evolution than the one used in ref .
we also compare the results obtained by using the gamma - ray information with the ones obtained by using the neutrino upper limit .
the telescope array measurement is considered to fit the uhecr flux assuming a pure proton component .
this is because , as mentioned above , the telescope array data are compatible with a proton composition at the highest energies . following refs .
@xcite the fits are performed for energies above @xmath11 ev .
it is assumed that the sources are uniformly distributed in the universe and that the injected spectrum follows a power law with an exponential cutoff , @xmath12 where @xmath13 is a normalization constant , @xmath14 is the spectral index , @xmath15 is the cutoff energy , and @xmath16 parametrize the source density and luminosity evolution of the injection with redshift @xmath8 .
the evolution function @xmath16 is not known , it depends on the types of sources responsible of the acceleration of the cosmic rays , which are still unknown .
the most common options found in the literature correspond to the evolution of the star formation rate ( sfr ) , the gamma - ray bursts ( grb ) , and the active galactic nuclei ( agn ) ( see for instance refs .
@xcite ) . in this work a broken power law of @xmath17
is assumed , @xmath18 where @xmath9 and @xmath19 are free parameters .
the break point at @xmath20 is motivated by the sfr , grb , and some special cases of agn evolution functions @xcite .
note that just the parameter @xmath9 is important for the fit of the spectrum because the contribution of the sources at @xmath2 is negligible for @xmath21 ev .
the evolution corresponding to @xmath2 is important for the generation of gamma rays and neutrinos .
the energy spectrum of uhecr at earth is calculated by using the program transportcr @xcite , which has been developed to solve numerically the transport equations that governs the propagation of the cosmic rays in the intergalactic medium .
one of the advantages of this approach is the decrease of the computational time compared with the one corresponding to the methods based on the monte carlo technique .
this makes it very useful for fitting the cosmic ray energy spectrum . besides nuclei
, the transportcr code propagates gamma rays and neutrinos produced in the interactions of them with the photon backgrounds .
the photopion production is based on the sophia code @xcite and there are several models of the ebl available
. models with fixed cutoff energy are considered .
it is possible to fit also @xmath15 as it is done in ref .
@xcite , but the computational time increases considerably . note that the processing time is larger in the case considered in this work because of the calculation of the propagation of gamma rays and neutrinos . for a given cutoff energy the proton flux at earth
is calculated for a discrete grid of the parameters @xmath14 , @xmath9 , and @xmath19 . as mentioned before
, the fit of the spectrum depends on the parameters @xmath14 and @xmath9 in the energy range under consideration , then the parameter @xmath19 just modifies the flux of secondary particles without affecting the fit .
that is why only some particular values of @xmath19 are considered .
the spectrum for any value of energy , @xmath14 , and @xmath9 is interpolated by using a trilinear interpolation algorithm , for fixed values of @xmath19 .
the fit of the telescope array spectrum is performed by minimizing the chi - square , which is given by , @xmath22 where @xmath23 is the interpolated spectrum , @xmath24 with @xmath25 the energy of the @xmath26-th data point and @xmath27 is a systematic shift in the energy scale ( see ref .
@xcite ) , @xmath28 is the flux measured corresponding to the @xmath26-th bin , and @xmath29 is the uncertainty in the determination of the flux also for the @xmath26-th bin .
the last term in the @xmath30 takes into account the systematic uncertainty on the determination of the energy , where @xmath31 ( see ref .
@xcite ) . following ref .
@xcite @xmath13 and @xmath27 are considered as nuisance parameters and then the profile likelihood technique @xcite is used to get rid of of them . in this technique
the nuisance parameters are replaced by their maximum likelihood estimators .
therefore , only @xmath9 and @xmath14 are the parameters of interest .
the ebl is still quite uncertain , principally because a direct measurement is subject to several foregrounds @xcite .
however , there are different models , available in the literature , for the spectrum as a function of the photon energy and redshift . in ref .
@xcite it was shown that the fit of the cosmic ray energy spectrum , assuming a pure proton composition , depends on the ebl model considered . moreover ,
the gamma - ray and neutrino spectra also depend on the ebl model assumed ( see refs .
@xcite and @xcite for gamma rays and neutrinos , respectively ) . in this work
we consider the ebl model of ref .
@xcite , which is commonly used to model the uhecr flux and also , following ref .
@xcite , the one of ref .
@xcite .
the top panel of fig .
[ kn21inf ] shows the fit of the telescope array energy spectrum considering the ebl model of ref .
@xcite , no cosmic ray emission above @xmath20 , i.e. @xmath32 , and @xmath33 ev . the telescope array spectrum is taken from @xcite and correspond to the spectrum obtained by the surface detectors .
the energy scale of the proton spectrum has been shifted according to the value @xmath27 obtained .
the best fit is obtained for @xmath34 , @xmath35 , and @xmath36 .
note that this results are compatible with the ones obtained in ref .
@xcite even though they do not use the penalty term in the @xmath30 .
the bottom panel of fig .
[ kn21inf ] shows the proton , gamma - ray , and neutrino spectra corresponding to the best fit .
also in this case , the energy scale of the proton spectrum has been shifted according to the value @xmath27 obtained .
the plot also shows the upper limit on the neutrino flux obtained by icecube at 90% cl @xcite .
the data points at low energies correspond to the spectrum of the egb observed by fermi - lat @xcite .
note that the best fit is compatible with the neutrino upper limit an also with the egb .
it is worth mentioning that the telescope array energy spectrum is extended to energies below @xmath11 ev with the combined spectrum taken from ref .
it can also be seen that the fit overshoots the data points below @xmath11 ev .
this low energy contribution can be suppressed due to the diffusion of the particles in the intergalactic magnetic field @xcite or by the presence of a low energy cutoff in the injection spectrum ( see ref .
@xcite for details ) .
( @xmath32 ) , and for @xmath33 ev . the shadowed areas correspond to 68.27% and 95.45% cl regions .
bottom panel : proton , gamma - ray , and neutrino spectra corresponding to the best fit of the telescope array data .
the filled circles correspond to the egb obtained by fermi - lat and the shadowed area corresponds to the rejection region at 90% cl for the upper limit on the neutrino flux obtained by icecube .
[ kn21inf],title="fig:",width=302 ] ( @xmath32 ) , and for @xmath33 ev .
the shadowed areas correspond to 68.27% and 95.45% cl regions .
bottom panel : proton , gamma - ray , and neutrino spectra corresponding to the best fit of the telescope array data .
the filled circles correspond to the egb obtained by fermi - lat and the shadowed area corresponds to the rejection region at 90% cl for the upper limit on the neutrino flux obtained by icecube .
[ kn21inf],title="fig:",width=302 ] as mentioned in the introduction , the egb measurement has important implications on proton models of the uhecr spectrum . by using the results obtained in ref .
@xcite , an upper limit on the integrated flux corresponding to the component of the egb spectrum that do not originate in point sources is obtained , @xmath37 at 90% cl . here the energy range considered for the integral goes from 50 gev up to 2 tev .
the method developed by feldman & cousins @xcite is used to obtain the upper limit ( see appendix [ ul ] for the details of the calculation ) . for each pair @xmath14 and @xmath9 considered to fit the telescope array data , after applying the profile likelihood method to the parameters @xmath13 and @xmath27 as described above , the integral of the gamma - ray energy spectrum is calculated and the region in the @xmath38 plane , defined by , @xmath39 is determined .
the top panel of fig .
[ regkn21 ] shows the best fit and the regions of 68.27% , 95.45% , and 99.73% cl for the fit in fig .
[ kn21inf ] . also shown
are the allowed regions inferred from the upper limits obtained form eq .
( [ iulreg ] ) for the case of gamma rays and from the condition @xmath40 , where @xmath41 is the upper limit on the flux obtained by icecube ( see bottom panel of fig .
[ kn21inf ] ) , for the case of neutrinos .
it can be seen from the plot that , while the best fit is compatible with the neutrino data , it is in tension with the upper limit obtained from the non - point sources component of the egb .
moreover , the best fit is also in tension with the upper limit obtained at 99% cl ( see appendix [ 99 ] ) . in this case
the upper limit obtained from gamma - ray data is more restrictive than the one coming from the neutrino data .
the bottom panel of the figure shows the result obtained in the same conditions as the ones corresponding to the top panel but for @xmath42 .
it can be seen that the best fit and the allowed regions are , as expected , unaltered , but in this case even the region corresponding to 99.73% cl is in tension with the gamma - ray upper limit . any value of the parameter @xmath19 larger than @xmath43 has associated a larger production of secondary gamma rays and neutrinos . therefore , the region corresponding to 99.73% cl of the fit is also in tension with at least the gamma - ray upper limit for any value of @xmath19 larger than 1.5 . .
the allowed regions corresponding to gamma - ray and neutrino observations are below the dashed and dotted curves , respectively .
bottom panel : best fit and confidence regions for @xmath42 .
the allowed regions corresponding to gamma - ray and neutrino observations are below the larger dashed and dotted curves and above the smaller ones in the bottom - left corner , respectively .
the ebl model of ref .
@xcite is considered and @xmath33 ev . [ regkn21],title="fig:",width=302 ] .
the allowed regions corresponding to gamma - ray and neutrino observations are below the dashed and dotted curves , respectively .
bottom panel : best fit and confidence regions for @xmath42 .
the allowed regions corresponding to gamma - ray and neutrino observations are below the larger dashed and dotted curves and above the smaller ones in the bottom - left corner , respectively .
the ebl model of ref .
@xcite is considered and @xmath33 ev . [ regkn21],title="fig:",width=302 ] from fig .
[ regkn21 ] it can be seen that for the case in which the uhecrs are generated in the redshift range from @xmath44 to @xmath20 , the gamma - ray upper limit imposes more restrictive conditions than the ones corresponding to the neutrino upper limit . however , when the sources produce uhecrs beyond @xmath20 the restrictions obtained from gamma - ray and neutrino observations are complementary , i.e. by using both , the neutrino and gamma - ray upper limits , it is possible to enlarge the rejection region .
the loss of restrictive power of the gamma - ray upper limit is due to the fact that the attenuation of gamma rays , owing to pair production with the ebl photons , that originate in @xmath45 is larger than the one corresponding to the gamma rays originated in @xmath46 .
this can be seen from fig .
[ specz12 ] where the contributions from @xmath47 $ ] and @xmath48 $ ] to the total spectra are shown .
the case considered corresponds to @xmath49 , @xmath50 , and @xmath42 , which is a point inside the 99.73% cl region of the fit for @xmath42 .
this point is also in the rejection region obtained from the gamma - ray data but in the allowed region obtained from the neutrino data .
in fact while the neutrino flux increases by a factor of @xmath51 at @xmath52 gev , when the component corresponding to @xmath48 $ ] is added , the integral of the gamma - ray spectrum increases in a factor of @xmath53 . , @xmath50 , and @xmath42 .
the curves labelled as 1 correspond to sources at @xmath47 $ ] and the ones labelled as 2 correspond to sources at @xmath48 $ ] . also shown are the total spectra obtained by adding the components 1 and 2.[specz12],width=302 ] figure [ in21inf ] shows the best fit and the associated gamma - ray and neutrino spectra for the ebl model of ref .
@xcite , for no uhecr emission above @xmath20 , and for @xmath33 ev .
the best fit is obtained for @xmath54 , @xmath55 , and @xmath56 .
although , the best fit parameters are different from the ones obtained for the other ebl model considered before , the differences between figs . [ in21inf ] and [ kn21inf ]
are small .
but for the ebl model of ref .
[ in21inf],title="fig:",width=302 ] but for the ebl model of ref .
[ in21inf],title="fig:",width=302 ] the top panel in fig .
[ regin21 ] shows the best fit parameters and the regions of 68.27% , 95.45% , and 99.73% cl for the fit of fig .
[ in21inf ] . also shown
are the allowed regions inferred from the gamma - ray and neutrino upper limits .
also in this case , the best fit is compatible with the neutrino upper limit but it is in tension with the gamma - ray upper limit at both 90% and 99% cl ( see appendix [ 99 ] ) . comparing with the top panel of fig .
[ regkn21 ] , it can be seen that , in this case , the upper limits coming from gamma - ray and neutrino data are less restrictive .
this is because the spectra of these secondary particles , predicted by using the second ebl model considered , are smaller in almost the entire energy range relevant for each type of particle . .
the allowed regions corresponding to gamma - rays and neutrino observations are below the dashed and dotted curves , respectively .
bottom panel : best fit and confidence regions for @xmath57 .
the allowed region from the gamma - ray upper limit is below the dashed curve .
the allowed region from the neutrino upper limit is below the large dotted curve and above the small dotted curve in the bottom - left corner .
the ebl model of ref .
@xcite is considered and @xmath33 ev . [ regin21],title="fig:",width=302 ] .
the allowed regions corresponding to gamma - rays and neutrino observations are below the dashed and dotted curves , respectively .
bottom panel : best fit and confidence regions for @xmath57 .
the allowed region from the gamma - ray upper limit is below the dashed curve .
the allowed region from the neutrino upper limit is below the large dotted curve and above the small dotted curve in the bottom - left corner .
the ebl model of ref .
@xcite is considered and @xmath33 ev . [ regin21],title="fig:",width=302 ] the bottom panel of fig .
[ regin21 ] shows the confidence regions of the fit and the allowed regions inferred from the gamma - ray and neutrino upper limits for @xmath57 . in this case
the 99.73% cl region of the fit is in tension with the neutrino upper limit and marginally with the gamma - ray upper limit .
also in this case the gamma - ray upper limit loss restrictive power due to the attenuation in the ebl , in such a way that , the neutrino upper limit is more efficient to reject all models whose points fall in the 99.73% cl region of the fit .
although @xmath58 ev is compatible with the telescope array data @xcite , the best fit value obtained in ref .
@xcite is @xmath59 ev with @xmath60 .
therefore , the same analysis done for @xmath58 ev is performed for this new value of the cutoff energy .
figure [ kn197inf ] shows the results obtained by using the ebl model of ref .
@xcite and for no uhecr emission beyond @xmath20 .
the best fit parameters obtained in this case are @xmath61 , @xmath62 , and @xmath63 .
the values obtained in ref .
@xcite are @xmath64 , @xmath65 , and @xmath66 which are close to the ones obtained in this work . in any case
, the differences can be explained by the different ebl model and propagation code used for the two calculations . comparing the bottom panel of fig .
[ kn197inf ] with the ones corresponding to figs .
[ kn21inf ] and [ in21inf ] it can be seen that for @xmath59 ev the production of gamma rays and neutrinos is smaller than for the case of @xmath58 ev .
this is due to the fact that @xmath59 ev is very close to the threshold energy of photopion production undergone by protons interacting with the cmb photons .
therefore , this channel for the generation of these secondary particles is quite suppressed . but for @xmath67 ev .
[ kn197inf],title="fig:",width=302 ] but for @xmath67 ev . [ kn197inf],title="fig:",width=302 ] the top panel of fig .
[ regkn197 ] shows the best fit parameters and the regions of 68.27% , 95.45% , and 99.73% cl for the fit of fig .
[ kn197inf ] . also shown
are the allowed regions inferred from the gamma - ray and neutrino upper limits . in this case ,
the best fit is compatible with the neutrino upper limit and it is marginally compatible with the gamma - ray upper limits .
as expected , in this case the gamma - ray and neutrino upper limits are less restrictive than for @xmath58 ev .
it is worth mentioning that the constraints imposed by the neutrino upper limit are less affected than the ones corresponding to the gamma rays , by the decrease of the cutoff energy .
this is due to the fact that , as can be seen from the bottom panel of fig .
[ kn197inf ] , the neutrino energy distribution becomes narrower than for the @xmath58 ev case and then the neutrino flux at the peak does not decrease considerably .
the bottom panel of fig .
[ regkn197 ] shows the results corresponding to @xmath42 in which it can be seen that , for this value , the region of 99.73% cl is in tension with the neutrino upper limit . in this case
the neutrino upper limit is more restrictive than the one corresponding to the gamma - ray observations . .
the allowed regions corresponding to gamma - rays and neutrino observations are below the larger dashed and dotted curves and above the smaller ones on the left - bottom corner , respectively .
bottom panel : best fit and confidence regions for @xmath42 .
the allowed region from the gamma - ray upper limit is below the large dashed curve and above the small dashed curve in the bottom - left corner .
the allowed region from the neutrino upper limit is below the dotted curve .
the ebl model of ref .
@xcite is considered and @xmath67 ev . [ regkn197],title="fig:",width=302 ] .
the allowed regions corresponding to gamma - rays and neutrino observations are below the larger dashed and dotted curves and above the smaller ones on the left - bottom corner , respectively .
bottom panel : best fit and confidence regions for @xmath42 .
the allowed region from the gamma - ray upper limit is below the large dashed curve and above the small dashed curve in the bottom - left corner .
the allowed region from the neutrino upper limit is below the dotted curve .
the ebl model of ref .
@xcite is considered and @xmath67 ev . [ regkn197],title="fig:",width=302 ] also for @xmath67 ev a similar behavior is obtained when the ebl model of ref .
@xcite is considered . in this case ,
the best fit parameters are @xmath68 , @xmath69 , and @xmath63 . for no uhecr emission above @xmath20
the best fit parameters fall in the allowed regions inferred from the gamma - ray and neutrino upper limits and the 99.73% cl region of the fit is in tension with the neutrino upper limit for @xmath70 .
in this work we have studied the constraints , imposed by the results recently obtained by the fermi - lat collaboration about the egb origin , on proton models of uhecrs . for that purpose ,
we have first calculated an upper limit to the component of the integrated egb flux , from 50 gev to 2 tev , that does not originate in point sources .
the obtained value at 90% cl is @xmath71 .
we have assumed that the uhecr emission as a function of redshift @xmath8 follows a broken power law in @xmath17 , with a breaking point at @xmath20 and an end point at @xmath72 .
we have found , by fitting the ta data , a very fast increase of the uhecr emission for @xmath47 $ ] , such that the index of the power law takes the values @xmath73 , depending on the cutoff energy and ebl model considered .
this result is consistent with previous work .
we have also found that for @xmath58 ev the best fit , corresponding to the case in which there is no uhecr emission beyond @xmath20 , is rejected at 99% cl by using the inferred upper limit from gamma - ray observations .
however , part of the 68.27% cl region of the fit is in the allowed region inferred from that upper limit . for @xmath59 ev the best fit is in the allowed regions obtained from the upper limits inferred from gamma - ray and neutrino observations . when the uhecr emission beyond @xmath20 is included the gamma - ray information becomes less restrictive .
this is due to the larger attenuation of the gamma - ray flux originated at @xmath2 in the energy interval considered for the integration .
the process responsible for this attenuation is pair production on the ebl . in this case , the gamma - ray and neutrino observations become complementary . by using both types of observations
we have found that the index of the evolution function for @xmath48 $ ] should be smaller than @xmath74 in order not to be in tension with the 99.73% cl region of the fit .
therefore , only proton models of the uhecrs with a much slower redshift evolution in the interval @xmath48 $ ] , compared with the the one corresponding to @xmath47 $ ] , are still compatible with the neutrino and gamma - ray upper limits . a more precise determination of the different egb components or a more restrictive neutrino upper limits are required to reject , at a given significance level , all proton models of uhecrs by using these types of analyses , which are independent of composition measurements .
* note added : * as our paper was completed another manuscript appeared on the arxiv , considering the impact of the egb measurements on proton models of uhecr , which contains complementary results to ours @xcite .
a. d. s. is a member of the carrera del investigador cientfico of conicet , argentina .
this work is supported by conicet pip 2011/360 , argentina .
the author thanks to o. kalashev and e. kido for their help with the transportcr program and the members of the pierre auger collaboration for useful discussions .
in this appendix the calculation of the upper limit on the egb flux component that does not originate in point sources , integrated from 50 gev to 2 tev , is presented . in ref .
@xcite it is shown that the integrated egb flux , corresponding to the energy range from 50 gev to 2 tev and originated in point sources is @xmath75 @xmath76s@xmath77sr@xmath77 .
the total integrated egb flux corresponding to the same energy range is calculated in ref .
@xcite and it is given by @xmath78 @xmath76s@xmath77sr@xmath77 . because @xmath79 has asymmetric errors the method developed in ref .
@xcite is used to find an approximated likelihood for the estimators @xmath80 and @xmath81 , which is given by , @xmath82 where @xmath83 @xmath76s@xmath77sr@xmath77 , @xmath84 @xmath76s@xmath77sr@xmath77 , and @xmath85 @xmath76s@xmath77sr@xmath77 .
here , @xmath86 where @xmath87 @xmath76s@xmath77sr@xmath77 and @xmath88 @xmath76s@xmath77sr@xmath77 .
the integrated egb flux that do not originate in point sources is obtained by subtracting the contribution of the point sources to the total one , i.e. @xmath89 .
therefore , introducing @xmath90 in eq .
( [ lall ] ) and applying the profile likelihood method @xcite to get rid of the nuisance parameter @xmath81 an approximated likelihood function for @xmath91 is obtained .
as expected , the minimum of @xmath92 is attained at @xmath93 @xmath76s@xmath77sr@xmath77 .
figure [ lik ] shows @xmath94 as a function of @xmath95 , where @xmath96 is the value of the likelihood at the maximum and @xmath97 @xmath76s@xmath77sr@xmath77 .
as can be seen from that figure the likelihood obtained is quite symmetric , this is due to the fact that the difference between @xmath98 and @xmath99 is not too big and that the likelihood corresponding to the point sources is combined with the one corresponding to the total egb integrated flux , which is symmetric . in order to simplify the calculations a gaussian approximation
is developed where the mean value is given by @xmath100@xmath76s@xmath77sr@xmath77 and @xmath101 @xmath76s@xmath77sr@xmath77 .
the gaussian approximation of the profile likelihood is also shown in the figure , it differs in less than @xmath102 in the two sigma region . in any case
, it is chosen such that the gaussian approximation is always smaller than the profile likelihood , which means that the upper limits obtained below are conservative . as a function of @xmath95 , with @xmath97 @xmath76s@xmath77sr@xmath77 .
the solid line corresponds to the profile likelihood and the dotted - dashed line to its gaussian approximation .
[ lik],width=302 ] the feldman & cousins @xcite method is applied in order to decide whether an upper limit or an interval containing the true value of @xmath103 , corresponding to a given cl , is obtained . by considering the following variable , @xmath104 the problem is reduced to the one corresponding to a gaussian distribution with unknown mean , @xmath105 , and standard deviation equal to 1 .
figure [ fc ] shows the confidence belts for 90% and 99% cl . , width=302 ] in this case , @xmath106 from fig .
[ fc ] it can be seen that , for this value , an upper limit is obtained for both , 90% and 99% cl .
the results are , @xmath107
figure [ reg2199 ] shows the rejection regions obtained from the upper limit on the integrated egb flux , corresponding to the energy range from 50 gev to 2 tev , that do not originate in point sources at 90 and 99% cl ( see appendix [ ul ] ) .
the cases considered correspond to @xmath33 ev and no uhecr emission beyond @xmath20 . from the plot
it can be seen that for both ebl models considered the best fit of the uhecr flux is rejected at 99% cl .
also shown are the rejection regions inferred from the neutrino upper limit . .
the allowed regions corresponding to gamma - rays and neutrino observations are below the dashed and dotted curves , respectively .
the curves corresponding to the gamma - ray data are labelled with the rejection probability .
the cutoff energy is @xmath33 ev . the ebl models considered are the ones in ref .
@xcite ( top panel ) and in ref .
@xcite ( bottom panel ) .
[ reg2199],title="fig:",width=302 ] .
the allowed regions corresponding to gamma - rays and neutrino observations are below the dashed and dotted curves , respectively .
the curves corresponding to the gamma - ray data are labelled with the rejection probability .
the cutoff energy is @xmath33 ev . the ebl models considered are the ones in ref .
@xcite ( top panel ) and in ref .
@xcite ( bottom panel ) .
[ reg2199],title="fig:",width=302 ] | the origin of ultra high energy cosmic rays ( uhecr ) is still unknown .
however , great progress has been achieved in past years due to the good quality and large statistics in experimental data collected by the current observatories .
the data of the pierre auger observatory show that the composition of the uhecrs becomes progressively lighter starting from @xmath0 ev up to @xmath1 ev and then , beyond that energy , it becomes increasingly heavier .
these analyses are subject to important systematic uncertainties due to the use of hadronic interaction models that extrapolate lower energy accelerator data to the highest energies .
although proton models of uhecrs are disfavored by these results , they can not be completely ruled out .
it is well known that the energy spectra of gamma rays and neutrinos , produced during propagation of these very energetic particles through the intergalactic medium , are a useful tool to constrain the spectrum models . in particular
, it has recently been shown that the neutrino upper limits obtained by icecube challenge the proton models at 95% cl . in this work
we study the constraints imposed by the extragalactic gamma - ray background , measured by fermi - lat , on proton models of uhecrs .
in particular , we make use of the extragalactic gamma - ray background flux , integrated from 50 gev to 2 tev , that originates in point sources , which has recently been obtained by the fermi - lat collaboration , in combination with the neutrino upper limits , to constrain the emission of uhecrs at high redshits ( @xmath2 ) , in the context of the proton models . |
in 70-th the notion of superspace is introduced in quantum field theory to make supersymmetry possible .
in addition to the usual dimensions ( coordinates ) @xmath0 , spinor fermionic ( anticommuting ) dimensions of space - time are introduced .
if supersymmetry is indeed discovered in the next generation of experiments at fermilab and cern within a decade or two ( see the review paper @xcite ) , this will amount to the discovery of the new dimensions .
there is another general problem in which the existence of additional spinor coordinates of space - time plays an important role . in 1952 ,
wick , wightman , and wigner @xcite show that the coherent superpositions of states with even and odd numbers of fermions are incompatible with the lorenz invariance and introduce the superselection rule , according to which such superpositions are physically impossible .
in actuality ( as is pointed out in @xcite ) , the superselection rule is the alternative to the existence of additional spinor coordinates .
the point is that the vector @xmath1 of any state with an odd number of fermions , being spinor of the odd rank , is multiplied by @xmath2 upon the rotations @xmath3 of the coordinate system through an angle of @xmath4 about any axis and upon the double time reversal @xmath5 .
the vectors @xmath6 of the states with an even number of fermions do not change under the @xmath3 and @xmath5 transformations .
for this reason , the existence of a coherent superposition @xmath7 where @xmath8 and @xmath9 are nonzero complex numbers , @xmath10 , implies the existence of a strange state @xmath11 which changes physically under the @xmath3 and @xmath5 transformations , because the corresponding change of the state vector does not amount to the appearance of a common phase factor .
if @xmath0 completely characterize the space - time , then the @xmath3 and @xmath5 transformations coincide with the identical transformation , which can change nothing physically .
the superselection rule is necessary in this case .
if , in addition to @xmath0 , the spinor coordinates exist , @xmath3 and @xmath5 are physically real transformations changing the sign of the additional coordinates . in this case , the superselection rule is not necessary .
thus , the proof of the possible existence of states @xmath11 corresponding to the coherent superpositions of states with even and odd numbers of fermions is simultaneously the proof of the reality of a superspace with additional spinor dimensions .
such proof is presented below .
namely , we show that states @xmath12 can be realized in a simple system with variable number of electrons which form , together with an environment , an isolated common system with a fixed number of electrons . in the case
considered , the interaction of the system with the environment can be described as an external field acting on the system .
this field has the spinor character ; i.e. , it changes sign under the @xmath3 and @xmath5 transformations .
all eigenstates of the hamiltonian of the system are coherent superpositions @xmath13 of the states with even and odd numbers of electrons , so at temperatures well below the characteristic energy difference , one ( ground ) of the states is automatically realized . irrespective of whether supersymmetry really exists or not , the existence of additional dimensions of space - time results in the existence of corresponding universal degrees of freedom for systems of fermions .
below , the nature of these degrees of freedom is elucidated for nonrelativistic systems .
we show that they correspond to the continuous change of the number of fermions in the system . in principle , there are two ways to change the average number @xmath14 of fermions continuously between two neighboring integral values @xmath15 and @xmath16 .
the states with nonintegral @xmath14 can be either impure states corresponding to incoherent mixtures of states @xmath15 and @xmath16 , or pure states @xmath11 considered above
. the first possibility would be the only one if the superselection rule is valid .
we show that the change of @xmath14 in a system from @xmath15 to @xmath16 through coherent superpositions @xmath13 can be ( and should be ) interpreted as `` the motion '' of the system `` as a whole '' along the additional fermionic dimensions of space - time . since the coherent change of @xmath14 is accompanied by the change of the spin of the system ,
the considered universal degrees of freedom include also the spin degrees of freedom .
we consider condensed matter systems such as metallic nanoparticles influenced by gate potentials at extremely low temperatures .
the nanoparticles behave like mesoscopic quantum dots ( mqd ) , i.e. , all ordinary spatial degrees of freedom of the particles are completely frozen out . near some critical values of the gate potentials ,
the particles possess electronic degrees of freedom which are not still frozen and which are described adequate in terms of the additional space - time coordinates .
we suggest experiments with mqds which generalize the experiment of nakamura et al .
@xcite on the observation of quantum coherence between the states with different ( but even in both cases ) numbers of electrons .
the implementation of these experiments will directly demonstrate the coherence between the states with even and odd numbers of electrons and , thereby , justify the reality of the new dimensions .
we are going to demonstrate that the superselection rule is not selfconsistent .
we show that states @xmath11 can be naturally realized in a simple realistic system with variable number of electrons because this system is governed by the hamiltonian whose eigenstates are all coherent superpositions of the states with even and odd numbers of electrons .
the idea ( see @xcite ) is as follows .
the number of electrons is a conserved quantity which is analogous , in this respect , to the momentum .
physical systems with hamiltonians whose eigenstates are all coherent superpositions of the states with different momenta are well known .
a particle in an external potential field depending on the particle coordinate is the simplest example .
in actuality , this particle is part of an isolated system consisting of the particle and a certain massive body , the interaction with which can be described as an external field acting on the particle . as known , this is justified only if certain conditions are fulfilled .
first , the states of a massive body must adiabatically adjust to the changes in the particle coordinate in order to prevent excitation of the body degrees of freedom .
second , all measurements must be made only with the particle .
no direct influence upon a massive body is possible .
let us consider single electron with fixed spin projection ( strong external magnetic field is applied along @xmath17-axis ) which can be localized in one of two quantum dots ( i and ii ) and can tunnel from one dot to another .
we are going to consider everything inside the dot i as system with variable number of electrons we are investigating ( analogous to the particle in the above example ) and everything inside the dot ii as environment ( analogous to a massive body ) .
this can be done by the following two steps .
first , we consider our initial system ( single electron ) as a part of a larger system with variable number of electrons consisting of both quantum dots whose hilbert space @xmath18 ( see fig.1 ) corresponds to the general superposition @xmath19 of all states @xmath20 with the numbers of electrons @xmath21 in the dot i and @xmath22 in the dot ii running through values @xmath23 independently .
let us introduce in @xmath18 the operators @xmath24 and @xmath25 of annihilation and creation of electrons in the dot i and the dot ii , respectively , so @xmath26 and @xmath27 . into the hilbert space @xmath28 the hilbert space @xmath29 of our initial system ( the straight line @xmath30 in fig.1 ) corresponds to the general superposition @xmath31 of states @xmath32 and @xmath33 .
the characteristic feature of the space @xmath29 is that the quantum number @xmath22 of the environment is determined unambiguously by the quantum number @xmath21 of the system : @xmath34 here the sign @xmath35 means , as in the book of dirac @xcite , that the corresponding equality is the following auxiliary condition on the state vector @xmath36 but not an operator identity in the whole space @xmath18 .
it is this property that allows the interaction of the system with the environment to be treated as an external field acting on the system .
this is a fermionic analogy to the adiabatic condition in the above example .
the second step is to project the hilbert space @xmath29 of the system into the hilbert space @xmath37 ( the straight line @xmath38 in fig.1 ) of the dot i considered as a separate system .
the corresponding state vectors @xmath39 are superpositions of states @xmath40 and @xmath41 . the projection is the result of the transformation @xmath42 of state vectors in the space @xmath18 with @xmath43 where @xmath44 .
we have @xmath45 the operator @xmath46 is not unitary in @xmath18 : @xmath47 but in @xmath29 , it satisfies the condition @xmath48 because @xmath49 the condition ( [ m ] ) gives the possibility to introduce the transformed hermitian operator @xmath50 acting in @xmath51 for each hermitian operator @xmath52 acting in @xmath29 in such a way that the matrix elements do not change .
in fact , let @xmath53 be two states in @xmath51 corresponding to two arbitrary states @xmath54 and @xmath55 in @xmath29 . according to ( [ eq10 ] ) , ( [ eq9 ] ) , and ( [ m ] ) we have @xmath56 the hamiltonian of single electron
can be represented in @xmath29 as @xmath57 where @xmath58 and @xmath59 are energies when electron is localized in the dot i and the dot ii , respectively , @xmath60 is the tunneling amplitude .
after simple calculations , we find quantities transformed to @xmath37 : @xmath61 @xmath62 @xmath63 where the condition @xmath64 in @xmath37 is used . finally , the hamiltonian of the system interacting with the environment is @xmath65 where @xmath66 .
we note that @xmath67 does not contain operators @xmath68 and @xmath69 of the environment .
the interaction of the system with the environment is described as an external field @xmath70 acting on the system .
the field @xmath70 , as well as the operators @xmath71 and @xmath72 , and other spinor quantities , change sign under the @xmath3 and @xmath5 transformations so that , for a given field value , hamiltonian ( [ eq16 ] ) is not invariant about these transformations . due to the presence of terms linear in electron operators , all eigenstates of the hamiltonian are coherent superpositions of the states with even and odd numbers of electrons
actually , ( [ eq16 ] ) is the hamiltonian of the two level system @xmath73
thus , to describe the above system correctly , we have to introduce additional spinor dimensions of space - time . assuming that the corresponding coordinates are the nonrelativistic limit of the coordinates considered in the field theory , we introduce ( as in @xcite,@xcite ) a pauli spinor @xmath74 where @xmath75 is a spin index .
additional coordinates @xmath74 are grassmann coordinates satisfying anticommutation relations @xmath76 whatever the actual superspace structure is in the relativistic case , this simplest possibility is quite general in the nonrelativistic limit . in the case considered above due to the presence of the strong magnetic field along @xmath17-axis , the system ( the dot i )
should be treated as `` moving '' along @xmath77 .
quantum mechanics with anticommuting coordinates is well known ( see for example @xcite ) .
the `` wave function '' @xmath78 is an analitical function of @xmath79 . due to the condition @xmath80 , the most general @xmath81 is @xmath82 where @xmath83 is the unit of grassmann algebra . by identifying @xmath84 and @xmath85
we see that states @xmath11 in ( [ eq17 ] ) are identical to ( [ eq010 ] ) .
the physical change of states @xmath11 under the @xmath3 and @xmath5 transformations is connected with the spinor nature of the physical coordinate @xmath79 .
operators @xmath72 and @xmath71 in the hamiltonian ( [ eq16 ] ) play the role of canonical coordinate @xmath86 and momentum @xmath87 corresponding to the additional dimension of space - time .
to elucidate the physical meaning of the degrees of freedom corresponding to the additional dimensions , it is very helpful to consider ground states of a mesoscopic quantum dot at different values of the gate potential or at different values of the electron chemical potential .
we consider a metal particle with a large but finite number @xmath15 of electrons , connected by tunnel junctions to macroscopic leads and ( or ) to other particles of the same type , and being influenced by a gate potential .
let us suppose that the temperature and all tunneling amplitudes are much smaller than the energy difference @xmath88 between the first excited and ground states of the particle at a given number of electrons , @xmath89 is fermi energy . under these conditions the metal particle behaves as a mesoscopic quantum dot ( mqd ) , i.e.
, it is a quasi - closed system in which all ordinary degrees of freedom associated with spatial motion of electrons are completely frozen . inasmuch as even at lowest experimentally possible temperature @xmath90 the number of electrons can not be larger than @xmath91
, we have to deal with metallic nanoparticles of the type obtained by ralph et al.@xcite .
thus , at a given @xmath15 the metal particle is in its ground state @xmath92 with the energy @xmath93 .
minimization of @xmath94 with respect to @xmath15 at a given gate potential @xmath46 determines the ground state value of @xmath95 .
we see that in mqd , a change in the number of electrons accompanying a change of the gate potential occurs as a result of first - order phase transitions between phases characterized by different integral values of @xmath15 ( see fig.2 ) .
the jumps of the number of electrons occur at critical values of the gate potential . vs gate potential @xmath46 at fixed @xmath96 it is important here to take into account the so - called parity effect , i.e. the fact that in fermi systems , the ground state energy @xmath93 calculated with an accuracy appropriate for a mesoscopic system contains in an explicit form the number of particles in the combination @xmath97 . in order to deal only with quasi - continuous functions at large @xmath15
, we should introduce two different functions @xmath98 and @xmath99 , individually for odd and even values of @xmath15 , extrapolated to the same value of @xmath15 .
the difference @xmath100 , which is usually positive , can be considered as a quantitative characteristics of the parity effect . due to the parity effect ,
the steps correspondong to even @xmath15 are longer than the steps corresponding to odd @xmath15 .
( see fig.2 ) .
with increasing @xmath101 the length of even steps increases , while the length of odd steps decreases until the odd steps disappear at all .
the jumps of the number of electrons at critical values of the gate potential become equal to 2 .
the phase transitions between states with neighboring even numbers of electrons take place at large @xmath101 .
a clear example of a system characterized by large @xmath101 is the so - called single - cooper - pair box ( see @xcite ) , i.e. a superconducting mqd in which all electrons form cooper - pairs and the energy of an additional single electron ( the superconducting energy gap ) is very large . to classify all possible ground states of a _
normal _ mqd at different values of the gate potential , let us put @xmath102 where the quasicontinuous number @xmath103 ( @xmath104 ! ) runs through all even numbers , while @xmath105 .
there are three states for each @xmath106 with @xmath107 , respectively , the combinations ( @xmath108 ) and ( @xmath109 ) being identical .
fig.2 shows the dependence @xmath110 at a given @xmath106 .
there are two critical values of the gate potential , @xmath111 and @xmath112 , corresponding to the conditions @xmath113 and @xmath114 respectively , where @xmath115 . as the parity effect increases ( with changing @xmath106 ) , the quantities @xmath116 and @xmath117 approach one another , so that for certain @xmath118 ( or @xmath119 ) determined by the equation @xmath120 a triple point may exist where all three states have the same energy .
we note that the triple point has actually been observed experimentally @xcite . for an even larger @xmath101 ( @xmath121 )
there is only one critical value , @xmath122 , where @xmath123 the jump of @xmath15 at @xmath124 is equal to 2 . at critical values of the gate potential ,
the ground state is degenerate .
according to the quantum superposition principle ( no superselection rule ! ) , there are infinite sets of ground states of the form @xmath125 @xmath126 and @xmath127 at @xmath128 and @xmath129 for @xmath130 , and at @xmath124 for @xmath131 , respectively , where @xmath132 .
the ground states of the form ( [ * ] ) , ( [ * * ] ) , and ( [ e ] ) are characterized by nonintegral averaged numbers of electrons @xmath133 , respectively .
one can say that these states correspond to `` phase coexistence '' regions ( vertical segments in fig.2 ) of first - order phase transitions taking place at @xmath128 , @xmath117 , and @xmath134 .
the `` phase coexistence '' of all three phases with @xmath135 , @xmath136 takes place at the triple point @xmath137 .
these ground states are characterized by two physical ( superconducting ! , see @xcite ) phases , @xmath138 and @xmath139 , and @xmath140 , @xmath141 , @xmath142 .
a more general consideration is the following . near critical values @xmath116 , @xmath117 , and @xmath134 of the gate potential
the mqd has some number of states which are close in energy to the ground state .
all other states have a much higher energy and can be neglected in low - frequency dynamics . in this sense
one can say that the mqd possesses degrees of freedom active at low temperatures and low frequencies ( much below @xmath143 ) .
active degrees of freedom are characterized by the hilbert spaces ( [ * ] ) , ( [ * * ] ) , ( [ e ] ) , or ( [ t ] ) near @xmath116 , @xmath117 , @xmath134 , or near the triple point , respectively . to take spin into account ,
let us suppose that the state @xmath144 and the state @xmath145 are spin - singlets , and the state with @xmath146 has a total spin 1/2 .
otherwise these states are `` incompletely frozen '' with respect to the bose degrees of freedom
. then in the most general case ( realized near the triple point ) , four states , @xmath144 , @xmath147 , and @xmath145 where @xmath75 is a spin index , are close in energy .
the hilbert space of active degrees of freedom corresponds to the general superposition @xmath148 where @xmath149 , @xmath150 , and @xmath151 are complex numbers .
active degrees of freedom near @xmath116 , @xmath117 , and @xmath134 , far from the triple point are described by ( [ g ] ) with @xmath152 , @xmath153 , and @xmath154 , respectively .
let us show that the active degrees of freedom described by ( [ g ] ) correspond to `` motion '' of the mqd along the spinor dimensions @xmath74 of superspace .
due to the anticommutation relations ( [ a ] ) , the most general wave function @xmath155 of the mqd is @xmath156 by identifying @xmath157 , @xmath158 , and @xmath159 , we see that the hilbert space ( [ g ] ) is identical to ( [ s ] ) .
the hamiltonian of the mqd is expressed in terms of the coordinates @xmath160 and momenta @xmath161 operators satisfying the canonical relations for fermi operators .
the operators @xmath162 and @xmath163 represent universal collective characteristics of any system of fermions under conditions such that the bose degrees of freedom are completely frozen . in exactly the same manner ,
the conventional coordinate and momentum operators describe the dynamics of a system with respect to a collective bose degree of freedom under conditions such that all other degrees of freedom are frozen .
the mqd is characterized by the following gauge invariant quantities @xmath164 where @xmath165 are the pauli matrices .
the operator @xmath21 corresponds to the parameter @xmath21 introduced above . the operator @xmath166 is the operator of the total spin of the mqd .
the operators @xmath162 and @xmath163 are therefore a generalization of the spin operators to the case where , together with @xmath0 , there exist additional dimensions described by @xmath74 .
the states ( [ * ] ) , ( [ e ] ) , and generally ( [ g ] ) are the stationary states of the mqd at critical values of the gate potential when the interaction of the mqd with the environment or with other mqds is neglected . manipulating by the gate potentials as functions of time in a system of two ( or more ) mqds one can realize conditions when even infinitesimal tunneling amplitudes cause observable tunneling transitions of electrons between different mqds .
this happens at the moments when the energies of different states of the total system coincide , the states ( [ * ] ) , ( [ e ] ) , and ( [ g ] ) playing the role of asymptotic states . the coherence of these states can be demonstrated by observations of interference phenomena ( see @xcite and below ) .
it is important to note that due to the relatively large dimensions of mqds one can measure the charge of a mqd without essential disturbance of neighbouring mqds .
this can be done ( as in the experiment of nakamura et al .
@xcite ) using a probe electrode connected to the mqd under study through a tunneling contact or ( as in the experiment of aassime et al .
@xcite ) using a probe electrometer based on a single - electron transistor .
let us consider the simplest experimental situation .
the role of quantum dots ( the dots i and ii ) in the discussion of section ii can be played by two mqds . the gate potentials should be close to corresponding critical values @xmath116 for both mqds .
the sum of the parameters @xmath21 ( determined by ( [ o ] ) ) should be equal to one . under these conditions
the tunneling of the additional ( with reference to the state with both @xmath21 equal to zero ) single electron between mqd i ( considered as a system ) and mqd ii ( considered as environment ) is the only active degree of freedom .
all other degrees of freedom correspond to much higher energies and can be neglected .
thus , to prove experimentally the existence of the new dimensions of space - time , one has to demonstrate the coherence of the superpositions ( [ eq17 ] ) for the two level system ( mqd i ) governed by the hamiltonian ( [ eq16 ] ) .
the corresponding time - dependent schroedinger equations are the system considered by nakamura et al .
@xcite to demonstrate the coherence between states with different , but even in both cases , numbers of electrons is also equivalent to a two level system described by the equations ( [ sch ] ) . in the case considered by nakamura et al .
the role of the system ( mqd i ) , the environment ( mqd ii ) , and the spinor field @xmath70 are played by a single - cooper - pair box , a macroscopic superconductor ( cooper - pair reservoir ) , and by the josephson coupling constant , respectively . below we consider two experiments .
the first experiment is the experiment of nakamura et al .
performed with our two - level system ( [ sch ] ) . before the initial time ( @xmath170 ) , the two - level system has been in the ground state with the gate potential of the mqd i such that @xmath171 .
accordingly , @xmath172 and @xmath173 . at @xmath170 ,
the gate potential rapidly changes to a value for which @xmath174 .
then , the potential remains constant for a time @xmath175 , after which it rapidly regains its initial value . on the time interval between @xmath170 and @xmath176 ,
the system obeys eqs.[sch ] with @xmath174 and initial conditions @xmath177 and @xmath178
. then @xmath179 and @xmath180 . at @xmath181 ,
one measures the everage charge of the system as a function of pulse duration @xmath175 .
the observed oscillations would indicate that the system coherently oscillates between the states with @xmath183 and @xmath184 on the time interval ( @xmath185 ) . as pointed out in section iii
, the system considered is an oscillator moving along the fermionic coordinate @xmath186 .
the nakamura - type experiment considered above can be modified by passing from the single - pulse to two - pulse technique .
as above , let the two - level system be at @xmath187 in the ground state @xmath188 and @xmath189 , @xmath171 .
the amplitude of the first gate - potential rectangular pulse is the same as above ( i.e. , corresponds to @xmath174 ) , but its duration is fixed at @xmath190 . immediately after the pulse at @xmath191 ,
the system is in the state with @xmath192 .
in the interval between @xmath193 and @xmath194 , the potential is equal to its initial value corresponding to @xmath171 . under this condition , the tunneling interaction of the system with environment can be ignored and it behaves as a closed system in its pure state characterized by the second gate - potential pulse with parameters of the first pulse is switched on at time @xmath197 . using eqs.([sch ] ) , one can see that , after completion of the second pulse at time @xmath198 ( @xmath199 because @xmath171 ) , the everage charge of the system is the observation of oscillations ( [ end ] ) as a function of time delay @xmath201 between the pulses would demonstrate that the state ( [ w ] ) is realized
. this would be direct experimental proof of the reality of a superspace with additional spinor dimensions .
99 j.h.schwarz and n.seiberg , rev.mod.phys .
* 71 * , s112 ( 1999 ) g.c.wick , a.s.wightman , and e.p.wigner , phys .
* 88 * , 101 ( 1952 ) a.f.andreev , jetp lett . *
68 * , 673 ( 1998 ) a.f.andreev , journ . of superconductivity
* 13 * , 805 ( 2000 ) ; physica b ( amsterdam ) * 280 * , 440 ( 2000 ) y.nakamura , yu.a.pashkin , and j.s.tsai , nature * 398 * , 786 ( 1999 ) a.f.andreev , jetp lett . * 74 * , 512 ( 2001 ) p.a.m.dirac , principles of quantum mechanics , oxford university press , 1958 , p.289 t.d.lee , particle physics and introduction to field theory , harwood academic publishers , 1981 , p.512 d.c.ralph , c.t.black , and m.tinkham , phys.rev.lett .
* 74 * , 3241 ( 1995 ) ; * 76 * , 688 ( 1996 ) ; * 78 * , 4087 ( 1997 ) m.t.tuominen , j.m.hergenrother , t.s.tighe , et al .
, phys.rev.lett .
* 69 * , 1997 ( 1992 ) a.aassime , g.johansson , g.wendin , et al .
, phys.rev.lett . * 86 * , 3376 ( 2001 ) | we note that the existence of physical states which are coherent superpositions of states with even and odd numbers of fermions means the existence , together with @xmath0 , of additional spinor dimensions of space - time .
a system with variable number of electrons is described in which such superpositions are realized .
experiments with mesoscopic condensed matter systems are suggested which generalize the experiment of nakamura et al . and may provide direct observation of such superpositions and , thereby , justify the reality of a superspace with additional spinor dimensions introduced in quantum field theory to account for supersymmetry .
the nature of additional dimensions of space - time is elucidated for nonrelativistic systems . _
p.l.kapitza institute for physical problems , russian academy of sciences , + kosygin str .
, 2 , 119334 moscow , russia _ |
quantum monte carlo ( qmc ) methods provide an important methodology for solving for the properties of interacting fermi systems . in auxiliary field techniques
@xcite , the partition function , @xmath1 $ ] is expressed as a path integral , for example , by discretizing the imaginary time @xmath0 into @xmath2 intervals of length @xmath3 , and separating the one body ( kinetic ) and two - body ( interaction ) terms .
the latter are then decoupled through the introduction of a hirsch - hubbard - stratonovich ( hhs ) field @xcite which reduces the problem to a quadratic form .
the fermion degrees of freedom can be integrated out analytically , leaving an expression for the partition function which is a sum over the possible configurations of the auxiliary field . for interacting lattice hamiltonians , such as the hubbard model ,
this field depends both upon the spatial site and on the imaginary time coordinate .
the sum over configurations is performed stochastically , for example , by suggesting local changes and accepting or rejecting with the metropolis algorithm .
the problem is challenging numerically because the summand is the determinant of a product of matrices , one for each fermion species .
the determinant is costly to evaluate , and can also become negative at low temperatures , which constitutes the fermion sign problem @xcite .
there are different ways to represent the matrices . in the determinant quantum monte carlo ( dqmc ) approach @xcite ,
the matrices have dimension equal to the number of spatial lattice sites @xmath4 .
the matrices are dense , and involve the product of @xmath2 sparse matrices .
the algorithm scaling , @xmath5 , arises from the need to update @xmath6 field variables at a cost of @xmath7 per update , where advantage is taken of an identity for the inverse and determinants of @xmath4-dimensional matrices which differ only by a rank - one change .
simulations with this method can now be done on many hundreds of spatial sites . in situations where particle - hole symmetry prevents a sign problem ,
for example , the half - filled hubbard hamiltonian , one can reach arbitrarily low temperatures .
dqmc simulations have proven the existence of long - range antiferromagnetic order in the two - dimensional half - filled hubbard model @xcite , as well as accurately determined the nature of the spectral function and thermodynamic properties at this density @xcite .
alternatively , in the algorithm developed by hirsch and fye ( hf ) @xcite for embedded - cluster problems , a larger , sparse matrix of dimension @xmath6 is considered .
the advantage of the hf - qmc approach is that the matrices are better conditioned ( no product of @xmath2 matrices is involved ) and also they remain positive to much lower temperatures ; the sign problem is far less severe in the hf - qmc method .
however , because determinants of larger matrices are involved , the hf - qmc algorithm scales as @xmath8 .
for this reason , hf - qmc has seen its most powerful applications within dynamical mean - field theory ( dmft ) @xcite and its cluster extensions , the dynamical cluster approximation ( dca ) @xcite , and the cellular dynamical mean - field theory ( cdmft ) @xcite for which @xmath4 is typically small . in effect
, dmft trades the large lattice sizes @xmath4 , and @xmath9 scaling of dqmc where spatial correlations can be explored , for the ability to reach much lower temperatures at general fillings at the cost of less real - space information , apart from that obtained from the mean field .
dmft also can directly access phase transitions which can only be inferred from finite - size scaling in dqmc . in this paper
, we describe a hybrid approach which combines some of the virtues of both dqmc and hf - qmc .
the key algorithmic improvement is a reduction in the @xmath10 hf - qmc scaling to linear in @xmath2 .
the importance is that this allows much larger @xmath4 to be considered . at the same time
, we demonstrate analytically ( and confirm numerically ) that the fermion sign problem in our hybrid algorithm is precisely the same as in hf - qmc , provided that the coupling to the host is fully taken into account .
thus , as in hf - qmc , we can reach low temperatures at quite general fillings .
our paper is organized as follows .
we first introduce the basic formalism , including a proof that the sign problem is unchanged from hf - qmc .
we then show results for various physical observables including the quasi - particle weight , local moment , and the green s function .
we demonstrate that the results of our algorithm converge to the same values as that of a well - developed and tested hf - qmc code .
we conclude with a comparison of the scaling properties of our new approach .
dmft , dca , and other cluster extensions such as the cdmft all map the lattice problem onto an effective cluster embedded in a self - consistently determined effective medium . here
, we will add additional sites to the cluster to emulate the effective medium @xcite .
the associated formalism will be sketched for the dmft and dca , but it is easily extendable to include cdmft .
the dca is a cluster mean - field theory which maps the original @xmath11dimensional lattice model onto a periodic cluster of size @xmath12 embedded in a self - consistent host .
this mapping is accomplished by replacing the green s function and interaction used to calculate irreducible quantities such as the self - energy ( @xmath13 ) by their coarse - grained analogs .
spatial correlations up to a range @xmath14 are treated explicitly , while those at longer length scales are described at the mean - field level . for details of the dca formalism and algorithm
, please see ref .
@xcite .
the dca loop converges when the cluster green s function equals the coarse - grained green s function , @xmath15 , @xmath16 where @xmath17 labels a cluster wave number , @xmath18 is the matsubara frequency , @xmath19 labels the lattice wave numbers in the wigner - seitz cell surrounding @xmath17 , and @xmath20 is the total number of lattice sites .
@xmath21 is the coarse - grained dispersion and @xmath22 is the single - particle hybridization between the dca cluster and its effective medium . here
, we consider the two - dimensional ( 2d ) single - band hubbard model . in order to employ dqmc as a cluster solver
, we define an effective cluster hamiltonian to preserve the coarse - grained green s function through the addition of host band degrees of freedom , which we label with @xmath23 .
@xmath24 the host band label , @xmath25 , runs from @xmath26 to @xmath27 .
@xmath28 is the dispersion for the @xmath23 band , @xmath29 is the coupling of the @xmath23 band to the @xmath30 band , @xmath31 is the strength of the interaction and @xmath32 is the number of spin-@xmath33 electrons on site @xmath34 . upon integration of the @xmath35band degrees of freedom ,
the correlated band green s function becomes @xmath36 where @xmath37 the parameters @xmath29 and @xmath28 are adjusted to fit the dca or dmft hybridization function @xmath38 . for this , we use marquardt s method @xcite to minimize the following merit function at each momentum point : @xmath39 we define the scaled deviation as @xmath40 where @xmath41 is the standard deviation of data .
the discretization of the bath degrees of freedom has been considered in dmft where exact diagonalization ( ed ) is used as the hamiltonian - based impurity solver @xcite .
extensions of this method to dynamical cluster mean - field theories have also been largely implemented to study variety of models such as the extended hubbard or multiband models @xcite .
the advantage of this method is that since ed is essentially exact , there is no systematic error beyond the discretization of the bath .
moreover , more complicated interactions than just the onsite coulomb can be easily included in the hamiltonian . however , the disadvantage of ed is that the hilbert space grows exponentially with the total size of the system , @xmath42 .
this greatly limits the size of the clusters that can be studied .
this is specially true since ( as we discuss below ) smaller clusters generally require a higher number of non - interacting bands to fully account for the coupling to the bath , and for larger clusters , e.g. , the 16-site cluster , even a very small @xmath43 will make ed inapplicable .
the average sign in the dqmc method is equivalent to the average sign in the hf - qmc method in the limit of infinite number of bath bands , @xmath44 . to prove this
, we use the path - integral formalism and write the partition function as @xmath45 \mathcal{d}[\gamma^*]e^{-s(\gamma,\gamma^*)}\ ] ] where @xmath46 $ ] denotes the functional integral , @xmath47 is the action , and @xmath48 and @xmath49 are grassmann variable vectors . equation ( [ eq : z - general ] ) can be approximated by @xmath50 \mathcal{d}[\gamma^ * ] e^{-s_0(\gamma,\gamma^ * ) } e^{-s_i(\gamma_c,\gamma_c^*)}\ ] ] where @xmath51 is the ( non)interacting part of the action and @xmath52 and @xmath53 represent the @xmath54band components . in eq .
( [ eq : z - appr ] ) , we have used hhs transformation to decouple the correlation in the interacting part of the action , @xmath55 here , @xmath56 , @xmath57 is the auxiliary field and @xmath58 is the time index so that @xmath59 .
the non - interacting part of the action has the following form : @xmath60\end{aligned}\ ] ] where @xmath61 is the non - interacting part of the hamiltonian and @xmath62 denotes both the spacial coordinate and the band index ( including the @xmath30 band ) .
equation ( [ eq : z - appr ] ) becomes exact in the limit of @xmath63 . by integrating out all the grassmann variables in eq .
( [ eq : z - appr ] ) , one obtains the following expression : @xmath64 \det [ g_{\downarrow}^{-1}]\ ] ] where @xmath65 is the green s function of size @xmath66 with @xmath67 . in the dqmc algorithm ,
@xmath68 $ ] is used as the sampling weight to complete the sum over the auxiliary field .
note that the action is off - diagonal in time , except for the first term of the non - interacting action which is equal to one along the diagonal [ see eq .
( [ eq : s0 ] ) ] .
therefore , @xmath69 is an off - diagonal sparse matrix with identity matrices along the diagonal and its determinant can be evaluated from a smaller matrix of size @xmath70 , using the following identity : @xmath71=\det[i+b_{\sigma , l}b_{\sigma , l-1}\dots b_{\sigma,2 } b_{\sigma,1}]\ ] ] where @xmath72 is the corresponding off - diagonal sub - matrix of @xmath69 at time slice @xmath58 .
the dqmc markov process proceeds by proposing changes in the hhs fields which are local in space and time , @xmath73 .
because of that , the ratio of the fermion determinants can be calculated directly from just the diagonal entry of the green s function .
similarly , the update of the green s function following an accepted move does not require a full @xmath74 matrix inversion , but can be done in @xmath75 operations .
more details about this algorithm can be found in ref .
@xcite . now suppose that instead of integrating out all the grassmann variables in eq .
( [ eq : z - appr ] ) , we integrate out only the ones associated with the non - interacting electron bands .
the partition function can then be written as @xmath76 \mathcal{d}[\gamma_c^ * ] e^{-s_c(\gamma_c,\gamma_c^*)}\ ] ] where @xmath77 in the above equation , @xmath78 is the non - interacting green s function on the cluster ( @xmath79 ) whose fourier transform to momentum and frequency space can be written as @xmath80 in the limit of an infinite number of non - interacting host bands , @xmath44 , the self - consistent dca hybridization function may be exactly represented by the analytic form of eq .
( [ eq : gamma ] ) , @xmath81 .
therefore , @xmath78 will be equal to the dca cluster - excluded green s function , @xmath82 . by integrating out the rest of grassmann variables in eq .
( [ eq : z - c ] ) , the partition function reads @xmath83 \det [ g_{c\downarrow}^{-1}]\ ] ] where @xmath84 is the dca cluster green s function of size @xmath6 . in hf - qmc , to complete the sum over the auxiliary field , @xmath85 is used as the sampling weight . unlike dqmc , where the inverse green s function is sparse
, here @xmath86 is a dense matrix with a dimension that grows with the number of time slices .
the hf - qmc markov process proceeds by proposing local changes in the hhs fields , @xmath73 .
the cost to propose a change , i.e. , to calculate the ratio of determinants [ eq . ( [ eq : z - hf ] ) ] , is low and does not depend upon @xmath2 or @xmath4 .
if a change is accepted , then the cluster green s function matrix @xmath87 must be updated .
it is possible to write this step as a rank - one matrix update .
however , since the inverse green s function matrix is dense , it is not possible to decompose it into @xmath88 blocks similar to what was done above with dqmc . by comparing eqs .
( [ eq : z - dqmc ] ) and ( [ eq : z - hf ] ) , one can write the following equation for a particular field configuration : @xmath89 \det [ g_{c\downarrow}^{-1}]= \det [ g_{\uparrow}^{-1 } ] \det [ g_{\downarrow}^{-1}].\ ] ] since @xmath90 is independent of fields , the ratio of sampling weights will be the same and therefore , the measured quantities , including the average sign , will have the same statistics in dqmc and hf - qmc algorithms .
) ] for @xmath91 , and @xmath92 versus matsubara frequency .
( b ) the corresponding scaled deviations of the fits [ eq . ( [ eq : corr ] ) ] versus temperature.,width=316 ]
we apply this method to the 2d hubbard model [ eq .
( [ eq : hamilt ] ) ] on a square lattice with nearest - neighbor hopping , @xmath93 , and show results for @xmath94 and the interaction equal to three quarters of the bandwidth ( @xmath95 ) at filling , @xmath96 , throughout this paper ; calculations at different doping regions and for interaction strength equal to the bandwidth lead to the same trends for the quantities discussed in this work .
the quality of the fit of the effective cluster hybridization function [ eq . ( [ eq : gamma ] ) ] to the dca or dmft hybridization function , @xmath22 , is improved by increasing the number of non - interacting bath bands . in fig .
[ fig : fitting](a ) , we show the imaginary part of @xmath97 and the corresponding data for @xmath98 from the fitting algorithm using different values of @xmath27 for a single impurity problem ( dmft ) .
the improved quality of the fit at a low temperature ( @xmath99 ) can be seen as @xmath27 increases from @xmath26 to @xmath92 .
we find that for a finite @xmath27 , the quality of the fit always decreases as the temperature is lowered .
this can be seen in fig .
[ fig : fitting](b ) where we show the scaled deviation of the fit [ eq .
( [ eq : corr ] ) ] for different values of @xmath27 as a function of temperature .
the hybridization function is poorly fit for @xmath100 even at high temperatures .
however , the scaled deviation is strongly reduced when @xmath27 increases . for a single impurity problem ( dmft ) .
we plot ( a ) the matsubara frequency quasi - particle fraction versus temperature , ( b ) the unscreened moment versus temperature , ( c ) the average sign versus inverse temperature and ( d ) the green s function at a low temperature versus imaginary time , calculated using hf - qmc and dqmc as impurity solvers . for comparison to exact results , a hf - qmc solution with very small @xmath101
is also presented . for dqmc
, we show results for @xmath102 , and @xmath92 . for a single - site problem ,
the average sign is exactly one in all cases .
the statistical error bars are smaller than the symbols and are not shown.,width=316 ] as the number of bath degrees of freedom increases , dqmc recovers the hf - qmc results for a single - site problem .
we find that a maximum of four bath bands are sufficient for the agreement of the two methods at temperatures as low as @xmath103 .
this convergence is shown in fig .
[ fig : nc1_n095 ] for @xmath104 where we plot the matsubara frequency quasi - particle weight ( @xmath105^{-1}$ ] ) , local moment ( @xmath106 ) and the green s function , calculated using hf - qmc and dqmc solvers .
to have an idea about the absolute errors , we have also included results from an exact solution , i.e. , hf - qmc with a very small @xmath101 ( @xmath107 ) .
we point out that the average fermion sign , shown in fig .
[ fig : nc1_n095](c ) , is equal to one , regardless of the bath in the single - site limit . for a @xmath108 cluster in the dca . in ( a ) and
( d ) , we plot the quasi - particle fraction at @xmath109 and the green s function at the origin , respectively
. results for @xmath110 can not be obtained due to a bad sign , even at relatively high temperatures.,width=316 ] the dqmc is a well - behaved cluster solver for the dca as the number of bath bands needed to recover the hf - qmc results decreases with increasing cluster size .
this can be understood from the suppression of the coupling between cluster and host degrees of freedom .
in fact , it was shown previously that the hybridization function in the dca is of order @xmath111 , where @xmath112 is the dimensionality @xcite . to illustrate that , we plot in fig .
[ fig : nc4_n095 ] the same quantities of fig .
[ fig : nc1_n095 ] using the same model parameters but now calculated on a @xmath108 cluster . for this cluster , the dqmc results show very good agreement with those of hf - qmc up to @xmath113 when @xmath114 .
as proven in the previous section , the average sign in dqmc converges to its hf - qmc value by increasing @xmath27 [ see fig . [
fig : nc4_n095](c ) ] .
we find that the sign shows a strong sensitivity to the quality of the hybridization function fit .
thus , when @xmath115 , results for @xmath100 can not be obtained due to a bad sign problem , even at relatively high temperatures . in figs .
[ fig : nc4_n095](a ) and [ fig : nc4_n095](d ) , we show the quasi - particle fraction at @xmath109 and the green s function at the origin in real space , respectively .
the dqmc cluster solver is best suited for larger cluster simulations where @xmath116 is sufficient to recover the hf - qmc results . as an example , we present results for a @xmath117 cluster in fig . [ fig : nc16_n086 ] .
we find excellent agreement between hf - qmc and dqmc calculations when @xmath116 . here
, the average sign falls more rapidly by decreasing temperature than that of the @xmath118 cluster [ see fig . [
fig : nc16_n086](c ) ] .
this limits the calculations for this cluster to @xmath119 in the optimally doped region .
however , as can be seen in fig .
[ fig : nc16_n086](c ) , the average sign is significantly improved from a finite - size dqmc calculation . for a @xmath120 cluster . for this cluster , the convergence of dqmc to hf - qmc
is achieved with @xmath116 . in ( c )
, we also show the average sign for a finite - size ( fs ) dqmc calculation on this cluster using the same model parameters.,width=316 ] as in hf - qmc , analytic continuation can be performed to calculate real - frequency quantities when dqmc is used as the cluster solver . as an example
, we have considered the case of fig .
[ fig : nc16_n086 ] and calculated the single - particle density of states ( dos ) using the maximum entropy method @xcite .
the results indicate that discretizing the bath degrees of freedom does not have a significant influence on the spectra .
a comparison between hf - qmc and dqmc dos has been presented in fig .
[ fig : doc ] where we find that there is a very good agreement between the two density of states in the low energy region .
however , there is a slight difference in the high - energy region which would presumably vanish by increasing @xmath121 . .
the solid ( dashed ) line shows the results for hf - qmc ( dqmc with @xmath116).,width=316 ]
as discussed in previous sections , the linear scaling of the dqmc algorithm with the number of time slices is the main advantage of this cluster solver over hf - qmc .
the updating process in hf - qmc , which is the most expensive step in this algorithm , scales like @xmath122 .
this is a results of @xmath123 changes in the field variable during each sweep and @xmath124 operations to update the green s function for each change , using a rank - one updating mechanism .
a similar argument applies to the scaling in the dqmc , except that it costs @xmath125 to update the inverse green s function after each change in the field variable . since the number of hhs fields and therefore , the number of such updates is proportional to @xmath2 , the overall scaling of updates in dqmc is linear in @xmath2 .
the scaling in the system size remains cubic as in other qmc methods and is a big advantage over ed which scales exponentially in the size . to show the linear behavior in @xmath2
, we plot the cpu time for updates versus @xmath2 on the @xmath120 cluster in fig . [ fig : update ] ( a ) .
first , we compare this to that of hf - qmc for the same model parameters and by setting @xmath126 . at this fixed @xmath0 ,
the product of matrices in dqmc is stable , which results in a perfectly linear scaling .
we find that the updating step in dqmc is up to three orders of magnitude faster than in hf - qmc for a large number of time slices ( @xmath127 ) . in more realistic simulations ,
increasing @xmath2 is a consequence of increasing @xmath0 to access low temperatures for a fixed order of systematic error ( constant @xmath101 ) @xcite . in this case
, we do not expect to see any change in the scaling of hf - qmc .
however , in dqmc , an orthogonalization step which scales as @xmath128 , has to be performed to avoid the round - off errors .
to show how the dqmc scaling changes , we also plot in fig . [ fig : update](a ) , the cpu time for dqmc with @xmath129 .
we see that the orthogonalization step introduces a quadratic term in @xmath2 with a coefficient which is two orders of magnitude smaller than the coefficient of the linear term [ see diamond symbols in fig .
[ fig : update](a ) ] .
this effect on the performance of the algorithm becomes slowly significant only when @xmath130 .
we point out that measuring the green s function in dqmc involves matrix multiplications of the same type as in the updating process , and therefore results in the scaling of the cpu time that is very similar to the one for the updates .
however , as can be inferred from fig .
[ fig : update](b ) , measurements generally take more time than updates and the quadratic term appears even in the case of constant @xmath0 .
the time for measuring the green s function in hf - qmc has more or less the same scaling as in dqmc , but is roughly an order of magnitude larger when @xmath131 . )
algorithms versus the number of time slices on a @xmath120 cluster .
all other quantities are kept constant .
the lines show power - law fits of the data .
the diamond symbols show the cpu time in dqmc with a constant @xmath3 ( decreasing temperature ) where orthogonalization is performed to stabilize the matrix multiplications.,width=316 ] ) algorithms versus the number of time slices on a @xmath120 cluster .
all other quantities are kept constant .
the lines show power - law fits of the data .
the diamond symbols show the cpu time in dqmc with a constant @xmath3 ( decreasing temperature ) where orthogonalization is performed to stabilize the matrix multiplications.,width=316 ]
in this paper we have shown that the use of dqmc as a cluster solver provides several order of magnitude speedup over the hf - qmc algorithm , with a sign problem which is well behaved ( identical to hf - qmc ) .
this improvement arises from a fundamental reduction in the scaling of the algorithm , from cubic in the inverse temperature , @xmath0 , to linear in @xmath0 ( with a small quadratic term arising from matrix orthogonalization to reduce round - off errors ) .
however , the hf - qmc approach itself has already been supplanted in many applications by `` continuous time '' qmc ( ctqmc ) algorithms @xcite .
we conclude this paper by addressing the relative strengths of the ctqmc technique and the new method presented here .
ctqmc eliminates the systematic error inherent in hf - qmc and dqmc , including the method presented here , by stochastically sampling the reducible feynman graphs of the partition function .
although the matrix sizes are generally smaller than in hf - qmc , the ctqmc algorithm also scales like the cube of the inverse temperature @xmath0 @xcite .
so , dqmc is generally much faster than ctqmc when applied to finite sized systems @xcite and also for the embedded cluster problems presented here , especially at low temperatures .
however , dqmc has the disadvantage of the introduction of systematic error .
these systematic errors in hf - qmc and dqmc may be eliminated by extrapolating the measured quantities in the time step squared , @xmath132 @xcite .
since the values of @xmath101 that are used in this extrapolation are not overly small , the linear in @xmath0 nature of the present algorithm makes for far more efficient calculations , especially at lower temperatures .
we have developed a dqmc cluster solver for the dmft , dca , or cdmft which scales linearly in the inverse temperature but has the same minus sign problem as hf - qmc .
formally , this is accomplished by defining an effective hamiltonian for the embedded - cluster problem which includes non - interacting bands for the host .
the additional hamiltonian parameters associated with the bath bands are adjusted to fit the cluster - host hybridization function .
we prove that when this fit becomes accurate , this dqmc algorithm recovers the same average sign as hf - qmc . using dca simulations of the two - dimensional single - band hubbard model ,
we demonstrate that as the number of bath bands increases , we recover the hf - qmc results , including the average sign .
the required number of bands is small , increases slightly with lowering temperature , and decreases with increasing cluster size .
we thank e. dazevedo , simone chiesa , and karlis mikelsons for stimulating conversations .
this work was funded by doe scidac project , grant no .
de - fc02 - 06er25792 which supports the development of multiscale many - body formalism and codes , including quest . e.k . and m.j .
were also funded by nsf grant no .
this research was enabled by allocation of advanced computing resources , supported by the national science foundation .
the computations were performed on lonestar at the texas advanced computing center ( tacc ) under account no .
tg - dmr070031n , and on glenn at the ohio supercomputer center under project no .
pes0467 .
`` quantum monte carlo methods in physics and chemistry '' nato science series , series c : mathematical and physical sciences
vol 525 , edited by m. p. nightingale and cyrus j. umrigar ( kluwer academic publishers , new york , 1998 ) m. h. hettler , a. n. tahvildar - zadeh , m. jarrell , t. pruschke , and h. r. krishnamurthy , phys .
b * 58 * , r7475 ( 1998 ) ; m. h. hettler , m. mukherjee , m. jarrell , and h. r. krishnamurthy , phys .
b * 61 * , 12739 ( 2000 ) ; m. jarrell , t. maier , c. huscroft , and s. moukouri , phys .
b * 64 * , 195130 ( 2001 ) . | dynamical mean - field theory and its cluster extensions provide a very useful approach for examining phase transitions in model hamiltonians , and , in combination with electronic structure theory , constitute powerful methods to treat strongly correlated materials .
the key advantage to the technique is that , unlike competing real - space methods , the sign problem is well controlled in the hirsch - fye ( hf ) quantum monte carlo used as an exact cluster solver .
however , an important computational bottleneck remains ; the hf method scales as the cube of the inverse temperature , @xmath0 .
this often makes simulations at low temperatures extremely challenging .
we present here a new method based on determinant quantum monte carlo which scales linearly in @xmath0 , with a quadratic term that comes in to play for the number of time slices larger than hundred , and demonstrate that the sign problem is identical to hf . |
although more than a quarter of a century has elapsed since the discovery of superconductivity in the quasi - one - dimensional ( quasi-1d ) organic charge - transfer ( bechgaard ) salts of @xmath3-type ( where @xmath4 stands for tetramethyltetraselenofulvalinium and @xmath5 being a strong electron acceptor or anion ) @xcite , many distinct properties of this material still remain a matter of debate . among those one may mention the pairing symmetry , the remarkable sensitivity of the critical temperature to irradiation @xcite , large upper critical field @xmath6 etcetera ( see , also ref . ) .
the irradiation destroys anion order , introducing thus non - magnetic damage into system that led abrikosov to suggest the possibility of triplet pairing @xcite in the organic salts .
nevertheless , the experimental evidence @xcite that the spin - susceptibility decreases at low magnetic fields in the superconducting ( sc ) state of the @xmath7 organic conductor , disfavors the triplet pairing mechanism and supports spin - singlet pairing .
low temperature properties of organic superconductors are known to be very sensitive to disorder .
alloying with anions , @xmath8-ray irradiation , or cooling rate controlled anion reorientation introduces non - magnetic randomness into the system , however leaving unchanged , to a large extent , the backbone structure and the unit cell of the organic superconductors .
there is a common agreement that disorder , when introduced by means of these experimental methods , must be characterized as non - magnetic . yet , it was shown @xcite to suppress the sc phase .
it is worth noting that the methods of generating non - magnetic disorder in layered organic superconductors are similar ( see , e.g. ref . ) to those in quasi-1d systems , and , therefore a similar mechanism of sc state suppression in these two classes of materials may exist .
studies of disorder effects on the superconducting phase have a long history .
the superconducting transition temperature @xmath0 for @xmath9-wave pairing has been shown to be insensitive to the scattering rate on non - magnetic impurities , which is known as the anderson theorem @xcite .
in contrast to non - magnetic impurities , paramagnetic impurities break time - reversal symmetry of the @xmath9-pairing , and suppress the sc - phase @xcite at some critical concentration of the impurities .
correlation between the paramagnetic impurities via the ruderman - kittel - kasuya - yosida ( rkky ) interaction yields a spin - glass phase below @xmath10 , which was shown @xcite to shift the superconducting transition point towards higher temperatures .
the anderson theorem is not applicable to unconventional superconductors with @xmath11-wave pairing symmetry @xcite .
a small concentration of non - magnetic impurities may destroy @xmath11-wave pairing , producing a finite lifetime for quasi - particles near the nodes in the gap @xcite .
inter - chain ( inter - layer ) pairings as well as intra - chain ( intra - layer ) pairings , occurring at different local critical temperatures in quasi-1d ( quasi-2d ) systems , yield also an inhomogeneous nodal order parameter @xcite , which affect considerably the upper critical magnetic field .
suppression of superconductivity in the presence of non - magnetic impurities can in general be realized by destroying either the modulus or the phase coherence of the order parameter .
interplay between superconductivity and anderson localization in a strongly disordered superconductor was shown @xcite to result in spatial inhomogeneity of the order parameter .
diffusive scattering of particles in the random field of non - magnetic impurities enhances coulomb repulsion @xcite , and consequently , reduces the amplitude of the order parameter .
mesoscopic fluctuations in a superconducting thin film were also shown @xcite to yield a spatial inhomogeneity of the order parameter .
effects of order parameter phase fluctuations on the superconducting transition temperature @xmath0 have been studied in refs .. it is well known that there is no sc phase transition in one - dimensional ( 1d ) and two - dimensional ( 2d ) systems @xcite , since strong fluctuations of the order parameter phase destroy off - diagonal long - range order ( odlro ) in a single sc wire and in an isolated sc film .
even a small interchain - coupling in clean quasi-1d superconductors restores however odlro together with a finite transition temperature .
the suppression of @xmath0 by strong phase fluctuations in clean quasi-1d superconductors was analyzed in ref .. classifying superconductors with small stiffness as bad metals , emery and kivelson have evaluated @xcite a critical temperature of phase ordering by formally dividing a clean bulk superconductor into small regions with well defined phase , and have shown considerable suppression of an sc - phase by phase fluctuations .
effects of disorder on phase fluctuations were however not considered in all of these papers .
distinct structural peculiarities of quasi-1d organic superconductors demand a special theory , which should take into account non - magnetic randomness as well as phase fluctuations .
indeed , the high purity of the organic superconductor backbone , even in an overall dirty limit , excludes spatial inhomogeneity of the order parameter modulus along the sc - wires .
this renders inapplicable the above - mentioned theories for pair breaking .
in contrast to these previous activities we study in this article a disorder - driven superconductor - normal metal phase transitions due to the competition of non - magnetic randomness and phase - fluctuations in quasi-1d superconductors .
we consider weakly linked quasi - one - dimensional superconductors with random josephson - couplings between pure one - dimensional ( 1d ) sc wires .
singlet pairing is assumed within the wires .
therefore , we assume that non - magnetic randomness does not affect the order parameter amplitude .
we demonstrate in this article that \(i ) non - magnetic randomness in the josephson - coupling destroys correlation of the phases between different chains in quasi-1d superconductors even in the classical phase fluctuation regime , \(ii ) non - magnetic randomness yields quantum critical behavior in addition .
a superconducting normal - metal phase transition occurs at @xmath2 increasing the strength of disorder , and \(iii ) a suppression of the sc phase occurs discontinuously as well both the classical and the quantum phase fluctuation regimes as a first - order phase transition when the disorder - strength reaches a critical value .
we derive parallel and perpendicular components of the penetration depth , @xmath12 and @xmath13 , and show that they diverge at different critical temperatures @xmath1 and @xmath0 , which correspond to pair - breaking in the wires and to phase coherence breaking between the sc wires , respectively .
the paper is organized as follows : in sec .
ii we study the interplay of randomness in the josephson energy with phase fluctuations inside the classical regime . although an arbitrary small interchain coupling in a clean quasi-1d superconductor stabilizes the odlro giving a finite transition temperature @xmath0 , we show in this section that non - magnetic disorder in the josephson energy suppresses @xmath0 discontinuously when the disorder - strength reaches a critical value . in sec .
iii , we investigate effects of randomness on the transverse rigidity and on @xmath0 in the quantum fluctuations regime .
we show in this section that a suppression of the sc phase is managed by two parameters characterizing the disorder - strength and a dynamical charging parameter @xmath14 in the system .
the quantum criticality at @xmath15 and the phase transition at finite temperature are described separately .
the transverse rigidity in the field of the phases is shown to vanish discontinuously .
the jump at the breakdown point decreases monotonically with increasing @xmath14 in the interval of @xmath16 , and vanishes for @xmath17 .
the breakdown point is pushed towards higher values of the disorder - strength as @xmath14 increases .
we show that , under a change of the disorder - strength , the critical temperature evolves similarly as the transverse rigidity at @xmath2 .
the analysis of diamagnetism in this section reveals completely different behavior for parallel and perpendicular components of the penetration depth . the penetration depth for a magnetic field parallel to the sc - wires
is shown to be defined by the phase - phase correlator between two neighboring wires , which behaves non - linearly in temperature and reveals a discontinuous dependence on disorder ; by contrast the perpendicular penetration depth does not depend on disorder and shows a conventional temperature dependence . in sec .
v we summarize results obtained in the paper .
the explicit evaluation of the phase - phase correlator is given in the appendix .
a quasi-1d superconductor is modeled as a system of one - dimensional wires , which are placed regularly and parallel to each other , forming for example a square lattice in the cross - section .
weak tunneling between the chains results in an open fermi surface for the normal metallic state , and yields also the josephson - coupling between nearest - neighboring chains in the superconducting state and strong anisotropy in kinetic properties . the free energy functional of a quasi-1d superconductor , which is weakly linked with josephson - coupling energy @xmath18 between nearest - neighbor chains , can be written in the presence of the magnetic field @xmath19 as @xmath20 { } \nonumber\\ & + & \xi_{\| } a^2_{\perp } \frac{({\bf b}({\bf r } ) - { \bf b}_{ext})^2}{8 \pi } \bigg \ } , \label{freeenergy}\end{aligned}\ ] ] where @xmath21 denotes the phase of the order parameter @xmath22 at a point with dimensionless coordinates @xmath23 , @xmath24 is the vector - potential , and @xmath25 the linear density of sc electrons with @xmath26 at @xmath27 , and @xmath28 with @xmath29 at @xmath30 .
dimensionless coordinates @xmath31 are introduced on the scale of the longitudinal @xmath32 and the transverse @xmath33 components of the coherence length , so @xmath34 . the last term in eq.([freeenergy ] ) describes the josephson - coupling with the coupling energy @xmath18 between the wires , which is minimal for a coherent tunneling @xmath35 of cooper pairs between neighboring wires .
fluctuations of the order parameter modulus can be neglected for pure superconductors @xcite far from @xmath1 , satisfying the condition @xmath36 , where @xmath1 is the mean - field critical temperature calculated for an isolated wire and @xmath37 is the number of bands in each chain @xcite .
therefore , the contributions to the free energy functional ( [ freeenergy ] ) , coming from the modulus of the order parameter @xmath38 , are omitted .
we assume the josephson energy @xmath18 to be a random parameter with gaussian distribution , centered at the mean value @xmath39 , given by @xmath40 the variance @xmath41 is taken as a measure of disorder strength in this coupling of nearest neighbor chains . employing the replica trick
one can calculate the average value of the free energy @xmath42 over disorder . as usual
we use @xmath43 and , in addition , express the n - th power of the partition function @xmath44 by means of replicated fields @xmath45 , @xmath46 , as ( for @xmath47 ) @xmath48\bigg ] + \frac{1}{2 } \sum_{\bf j , g}\bigg[\frac{n_s \xi_{\|}w}{t}\times \hspace{+1.0 cm } \nonumber\\ \times \int dz\sum_{a } [ 1-\cos ( \varphi_{\bf j}^a - \varphi_{\bf j+g}^a)]\bigg]^2 \bigg \ } , \hspace{+3.0cm}\end{aligned}\ ] ] the quadratic cosine term is linearized with the help of a hubbard - stratonovich transformation by introducing an auxiliary field @xmath49 . as a result , the sum over the replica variable @xmath50 is factorized and the replica limit can be performed , yielding for the averaged free energy @xmath51
\bigg\}. \label{eff - energy}\end{aligned}\ ] ] the average value of a given functional @xmath52 , e.g. @xmath53 or @xmath54 , can be obtained according to the relation @xmath55 by adding the source term @xmath56 to the free energy functional , which yields for the correlator @xmath57 where the double bracket @xmath58 means averaging over thermodynamic fluctuations and over randomness . in order to estimate an asymptotic behavior of the correlator , e.g. @xmath59 we write the integrand of eq.([av ] ) as @xmath60 , and apply the stationary - phase approximation to determine an extremal value of the auxiliary field @xmath61 minimizing the function @xmath62 . the minimal value of @xmath49 is obtained to be @xmath63 the constant @xmath64 on the exponent can be estimated to be equal to @xmath65 for the organic superconductors with @xmath66 being the fermi energy , which ensures a sharply peaked saddle point of the integrand .
the thermodynamic averages in the expression of @xmath67 are taken with the free energy functional , given by eq.([eff - energy ] ) , at the saddle point @xmath68 .
so , a contribution of the non - magnetic randomness to the effective free - energy functional is proportional to the variance of the phase correlator , which gives an idea on the form of the disorder - dependent term in the effective functional .
note also that the saddle point for the averaged free energy @xmath69 is given as @xmath70 \rangle$ ] , where @xmath71 is the value of the free energy at the saddle point . the critical temperature for the quasi-1d superconductors can now be found from eq.([av]),written for @xmath72 by using the self - consistent mean - field method @xcite , which consists in replacing the phase correlations of the cosine term by @xmath73 \nonumber\\ & \longrightarrow & % \to e_{\perp } [ 1 - \langle \langle \cos(\varphi ) \rangle \rangle \cos ( \varphi ( z))],\end{aligned}\ ] ] where @xmath74 . for a clean system @xmath75
was chosen @xcite to be equal to @xmath76 .
for the disordered superconductor we choose @xmath77 .
the functional integral over the phases in eq.([av ] ) can not yet be evaluated , even after this simplification .
taking advantage of the smallness of @xmath78 near @xmath0 , we expand both the numerator and the denominator of the integrand of eq.([av ] ) , written for @xmath79 , in this parameter .
the thermodynamic averages become pure one - dimensional after this expansion , which can be taken easily , yielding a power series of @xmath80 for the integrand . therefore
, the integration over @xmath80 is immediately performed .
since all higher order in @xmath81 terms of the expansion vanish at @xmath82 , we get the equation for @xmath83 @xmath84 where @xmath85 is the coordination number .
the phase - phase correlator in eq.([tc ] ) is calculated in the clean limit of the @xmath86 free energy functional , obtained from eq.([freeenergy ] ) by setting @xmath87 , which returns ( see for example ref .
@xcite ) @xmath88 where @xmath89 .
introducing a dimensionless @xmath0-shift by @xmath90 with @xmath66 being the fermi energy , and a dimensionless disorder parameter @xmath91 eqs.([tc ] ) , ( [ 1dcorrelator ] ) yield @xmath92 the full solution of eq.([tc3 ] ) has three roots , among which the physical one is confined to the finite @xmath93-range as shown by the bold line in fig.[tc - classic ] .
a ( physical ) solution exists thus only within the finite range between clean limit ( _ cl _ ) and dirty limit ( _ dl _ ) .
one may expand and control this physical solution ( of the cubic equation ) in the weak disorder regime ( small @xmath93 ) , where the @xmath0-shift obeys @xmath94 showing that @xmath0 decreases with increasing randomness like @xmath95 .
for a pure system eq.([criticalt ] ) gives the dependence @xmath96 , in agreement with efetov and larkin in ref . .
this expression shows that even a small interchain - coupling sustains odlro in the system and , consequently , the critical temperature increases with @xmath97 . on the other hand
, the competing destructive effect of disorder reduces @xmath0 due to `` melting '' of the order parameter phase coherence between neighboring chains . according to the ( physical ) solution of eq.([tc3 ] ) , the critical temperature decreases monotonically with increasing @xmath93 but finite @xmath0 are confined to the interval @xmath98 .
the sc - phase becomes fully suppressed for disorder - strengths @xmath41 exceeding a critical disorder - value @xmath99 given by @xmath100 beyond which the system is in a normal metallic phase ( for @xmath101 ) .
the critical temperature drops to zero at @xmath102 with a jump of size @xmath103 thus the sc - normal metal phase transition appears as a @xmath104st - order transition . in order to describe the behavior of @xmath105 near the disorder limit _ dl _ ( see fig.[tc - classic ] )
we expand around @xmath106 , in terms of small ( nonnegative ) @xmath107 and @xmath108 , which gives @xmath109 . near the dirty limit
, the @xmath0-variation has an infinite slope ( see also fig.1 ) .
this can be reexpressed in terms of the physical parameters @xmath110 and the variance @xmath41 of the josephson - coupling , by reinserting eqs.([q - eq],[criticalt ] ) , as @xmath111 in the vicinity of the breakdown point @xmath112 thus the critical temperature decreases with disorder almost linearly but , approaching the dirty limit @xmath113 , it finally turns into a ( nonanalytic ) square - root behavior close to the breakdown point . in the absence of the non - magnetic disorder , even arbitrarily small josephson - coupling between the chains stabilizes the odlro and gives a non - zero critical temperature .
however , the sc phase with finite josephson - coupling can be destroyed by increasing the strength of non - magnetic disorder .
the self - consistent mean - field method , applied above for the classical phase - fluctuation regime , expressed the @xmath0-equation in terms of the pure @xmath104d phase - correlator ( [ 1dcorrelator ] ) , neglecting in this respect the josephson - coupling between neighboring chains .
we shall now improve the calculation of the phase - correlator by taking into account the transverse rigidity of the system , which provides a more realistic determination of the transition temperature in the quantum fluctuation regime .
our calculations are carried out in the hamilton formalism for convenience , yet the problem can be formulated in the path integral language @xcite as well .
let us start from the lagrangian , again taking @xmath47 for simplicity , @xmath114 ^ 2 - f_{st}^{eff}\{\varphi \ } , \label{lagrangian}\ ] ] where @xmath115 denotes the time derivative of the phase .
the dynamical term in the lagrangian can be interpreted as the electrostatic energy of charged wires @xcite @xmath116 generated according to the first josephson equation @xmath117 , and @xmath118 are the specific coefficients of electrostatic induction . rewriting the electrostatic energy @xmath119 in terms of the time - derivative of phases , the fourier transform @xmath120 of the coefficients @xmath121 , has the physical meaning of a compressibility . in eq.([lagrangian ] ) we neglect dispersion in the compressibility and assume @xmath122 this approximation is equivalent to a replacement of the electrostatic energy ( [ elec ] ) by @xmath123 .
the parameter @xmath124 can be calculated @xcite in the presence of coulomb screening for a small born parameter @xmath125 , which results in @xmath126^{-1}. \label{compres}\ ] ] there , @xmath127 is the unscreened compressibility , @xmath128 denotes the longitudinal velocity of an electron on the fermi surface averaged over @xmath37 subbands , while @xmath129 and @xmath11 stand for the interchain - distance and the diameter of a superconducting wire , respectively . in order to get @xmath130 we average the free energy over disorder and apply a mean - field approximation , corresponding to the replacement @xmath131[1-\cos ( \varphi_{\bf j}(z ' ) - \varphi_{\bf j+g}(z'))]$ ] by @xmath132\rangle \rangle \int dz [ 1-\cos ( \varphi_{\bf j}(z ) - \varphi_{\bf j+g}(z))].$ ] the hamiltonian , expressed through the phases @xmath133 and canonical conjugate momenta @xmath134 , becomes @xmath135\biggr ] \bigg\ } , \label{hamilton}\end{aligned}\ ] ] where @xmath136 , while @xmath137 is given by @xmath138\rangle \rangle \biggr ] \label{rigidity - dis}\ ] ] and represents the dimensionless anisotropy - parameter or the transverse rigidity of the system ; @xmath139 in eq.([rigidity - dis ] ) is transverse rigidity of the pure system @xmath140 the phase dynamics in the classical limit can be obtained from the euler - lagrange equation , which is described by a set of coupled sine - gordon type non - linear equations @xmath141 where @xmath142 is a characteristic scale of frequency , and @xmath143 we express @xmath142 as @xmath144 , where @xmath145 the parameter @xmath14 is the essential parameter of the theory , which can be written , using eq.([compres ] ) , as @xmath146^{1/2}.\ ] ] the system of equations ( [ dynamics ] ) is linearized for small @xmath133 and its fourier - transformation becomes diagonal @xmath147 the eigenfrequency of oscillations @xmath148 is given in the harmonic approximation as @xmath149^{1/2}. \label{frequency}\ ] ] these equations describe the frequency of low - lying plasmon - mode of the system .
the quantum description is realized by expressing @xmath150 and @xmath151 as a linear superposition of bose operators @xmath152 and @xmath153 , @xmath154 = 1 $ ] , by @xmath155 if we expand the cosine term in eq.([hamilton ] ) up to the quadratic term and express the phases @xmath156 and the conjugate momentum @xmath151 in terms of creation and annihilation operators , we get the hamiltonian in the harmonic approximation as @xmath157 , \label{hamiltonian}\ ] ] where the energy spectrum is defined by eq.([frequency ] ) . in order to take into account the quantum effects in the hamiltonian
, we have to express the cosine term in eq.([hamilton ] ) in a normal ordering before expanding over @xmath153 and @xmath152 @xmath158 = \delta_{cl}^2 \biggl [ 1 - \nonumber\\ & & \hspace{-1cm}1-\frac12 e^{- s_{\alpha}({\bf g } , 0)}\left ( e^{\sum_{\bf q } a_{\bf q } b_{\bf q}^{\dag } } e^{- \sum_{\bf q } a_{\bf q}^{\star } b_{\bf q } } + % e^{- \sum_{\bf q } a_{\bf q } b_{\bf q}^{\dag } } % e^{\sum_{\bf q } a_{\bf q}^{\star } b_{\bf q } } h.c.\right ) , \label{operator}\end{aligned}\ ] ] where @xmath159 the prefactor @xmath160 originates in the commutation relation between @xmath161 and @xmath152 , and by taking into account the baker - campbell - hausdorff relation @xmath162 \}$ ] .
furthermore @xmath163 where @xmath164 is the number of unit cells per volume .
it is clear from eq.([operator ] ) that the physical meaning of @xmath165 is an average of @xmath166 over the ground state at @xmath15 . in the framework of the self - consistent phonon approximation ( scpa ) , we expand the josephson term ( [ operator ] ) in powers of the creation and annihilation operators , @xmath153 and @xmath152 repsectively . expressing the leading ( harmonic ) part of the hamiltonian ( [ hamilton ] ) in terms of the particle number operator @xmath167
, we obtain again a harmonic hamiltonian as in eq .
( [ hamiltonian ] ) , @xmath168 , but with the different oscillation frequency @xmath169^{1/2 } , \label{frequency0}\ ] ] thus the application of the scpa results in a re - normalization of the parameter of anisotropy @xmath137 in the frequency of oscillation @xmath170 by means of the phase - phase correlator as @xmath171 we note that in order to preserve a maximal coherence of phases at @xmath172 in the josephson term @xmath173 the latter is replaced in the framework of the scpa by @xmath174 , which corresponds to shifting of the energy origin .
indeed , the josephson term in the initial expression of the hamilton function was introduced in a such way that it becomes zero for a maximal coherence of phases @xmath172 .
zero - point fluctuations at @xmath2 in the quantum case destroy the phase coherence and increase the josephson energy . by shifting @xmath175
we again reach a minimal josephson energy at @xmath2 in the quantum case too .
similar shifting was done also in eq.([rigidity - dis ] ) . in the expansion of the exponential operator of eq.([operator ] )
we select all diagonal terms , which can be expressed in terms of the bosonic particle number operator @xmath167 @xcite .
this yields @xmath176 absorbing now the product in eq.([expeqn ] ) into the exponential form , and neglecting all higher orders in @xmath177 beyond the leading term @xmath177 ( for justification see below [ ] ) , we get @xmath178 thus , after this step of calculations we still restrict ourself with harmonic approximation , describing the system by the hamiltonian @xmath179 like ( [ hamiltonian ] ) where the transverse rigidity @xmath180 in the frequency @xmath181 is renormalized as @xmath182 where @xmath183 the trace over the diagonal part of the phase - phase correlator within the harmonic approximation replaces the bosonic filling number operator @xmath184 by the planck distribution function for phonons with energies of @xmath185 as @xmath186 , yielding the following expression for the @xmath187 @xmath188 the correlator @xmath189 and its @xmath2-limit , as given by eqs.([correlatort ] ) and ( [ s ] ) respectively , are evaluated explicitly in the appendix .
the zero - temperature behavior of the system is analyzed by means of the phase - phase correlator @xmath190 , the explicit expression for which is given by eq.([s(g,0)3 ] ) in the appendix .
expressing the phase - phase correlator @xmath190 in terms of @xmath191 gives @xmath192 , which implies that even a small interchain - coupling stabilizes odlro in the system , hence also a finite @xmath193 phase transition should exist . in order to get an explicit expression for the dependence of @xmath194 on @xmath195 and on disorder
, we have to solve the equation @xmath196 together with eq.([rigidity - dis ] ) for @xmath137 .
thus the equation for the reduced transverse rigidity @xmath197 , where @xmath198 is the renormalized transverse rigidity for the clean system at @xmath15 , assumes the form @xmath199 where the quantum parameter of randomness @xmath200 reads @xmath201 @xmath202 in eq.([qofw ] ) is a constant @xmath203 .
although eqs.([tc3 ] ) and ( [ delta - qu ] ) are written for two different characteristic parameters of the system , it is easy to see that the equation for @xmath204 , if we neglect the quantum effects at @xmath205 , coincides with eq.([tc3 ] ) written for @xmath206 .
eq.([delta - qu ] ) can be solved approximately for moderately weak disorder , yielding the following expression for @xmath207 @xmath208^{\frac{1}{1 - \alpha}}. \label{delta - qu0}\ ] ] hence , the evaluation of the phase - phase correlator @xmath209 in the presence of disorder yields the result @xmath210^{\frac{\alpha}{1 - \alpha}}. \label{s(g,0)f}\end{aligned}\ ] ] in the absence of the disorder , i.e. for @xmath211 , we retrieve the phase - phase correlator @xmath212 as obtained by efetov and larkin @xcite for pure quasi-1d superconductors .
the critical temperature for a quasi-1d superconductor , according to efetov and larkin @xcite , is defined by @xmath213 for pure superconductors and by @xmath214 in our case for dirty superconductors .
this relation with eq.([s(g,0)f ] ) shows that @xmath215 decreases nonlinearly with disorder . the numerical solution of eq.([delta - qu ] )
is depicted in fig.[tc - quantum ] .
the reduced @xmath2 transverse rigidity @xmath216 is shown to decrease with increasing disorder for ( fixed ) @xmath217 , and suddenly drops to zero at the critical disorder strength @xmath218 .
hence the quantum critical behavior corresponds to a first order phase transition .
fig.[tc - quantum ] shows how the breakdown point shifts with increasing @xmath14 to higher values of the randomness , and the jump vanishes as @xmath219 .
eq.([delta - qu ] ) becomes linear for @xmath220 and gives , by inferring the @xmath221-relation from eq.([qofw ] ) , @xmath222.\ ] ] here , the transverse rigidity @xmath223 decreases linearly with increasing @xmath41 and vanishes at @xmath224 the quantum critical behavior in the model is however controlled by two parameters , the strength of randomness @xmath221 and the parameter of quantum dynamics @xmath14 . for @xmath217 , the superconductor - normal metal phase transition at @xmath2 is always discontinuous , and only turns into second - order at @xmath220 .
let us now study the _ finite @xmath193 behavior _ of the transverse rigidity .
the phase transition in a quasi-1d superconductor occurs at some temperature @xmath225 when the transverse rigidity in the ensemble of phases @xmath226 vanishes , which results in `` melting '' of the phase coherence .
the energy spectrum @xmath227 of the collective excitations is reorganized and the transverse @xmath228-dependent part of @xmath227 vanishes at @xmath225 , i.e. symmetry breaking occurs in the bosonic excitation at @xmath225 . inserting the solution of eq.([s ] ) for @xmath229 into @xmath230 and using eq.([delta - qut ] ) , we obtain @xmath231 where a new temperature scale is introduced by means of @xmath232 . in terms of @xmath233 and @xmath234 , eq.([deltatc ] ) assumes the form @xmath235 , which has a non - zero solution only for @xmath236 .
the finite solution of this equation vanishes discontinuously at @xmath237 , giving the following value for @xmath0 @xmath238 the magnitude of the jump in @xmath239 is @xmath240 , and hence the phase transition is of first - order .
the similar behavior has been found also in the planar rotor model @xcite in the absence of disorder .
the dependence of @xmath0 on the disorder is determined by the zero - temperature transverse rigidity @xmath207 , the behavior of which is depicted in fig .
[ tc - quantum ] .
therefore , for arbitrary @xmath217 the critical temperature decreases monotonically with increasing randomness and drops to zero at the critical disorder strength @xmath241 .
the variation of the critical temperature versus the residual resistivity of the organic superconductor @xmath242 has been explored by joo et al . in ref .. the experimental data are read off from ref . and copied into fig.[exper ] in order to provide a close comparison with a theoretical fit - curve as obtained from eq.([fig3-curve ] ) of the present theory . in this first approach and within a moderate accuracy , the data as published in the paper of joo et al .
, eventually appear to find an explanation by our theory . in order to establish the link between experiment and theory ,
the following argument is exploited : the substitution of @xmath243 anions by @xmath244 in the relaxed ( r ) samples increases the residual resistivity , which is proportional to the inverse lifetime @xmath245 of electron as well as to the disorder strength @xmath41 ( as used in the present theory ) . for the relaxed samples of ref .
, referred to as the r - sample(s ) , the doping concentration @xmath8 varies in the interval of @xmath246 under slow cooling from the clean sample to the nominal concentration @xmath247 .
slow cooling of @xmath248-samples assures a uniform orientation of the anions along the stacking axis , whereas fast cooling in the ( quenched ) @xmath249-samples introduces strong orientational disorder and increases the residual resistivity .
@xmath0 decreases quasi - linearly with increasing disorder ( or the residual resistivity ) in the large interval of the randomness . around the breakdown point
the dependence of @xmath0 on randomness is non - linear .
the critical doping concentration , corresponding to the breakdown of the superconducting state , grows with the quantum charging effect in the system .
substitution of @xmath243 anions by @xmath250 seems to increase the quantum charging parameter @xmath14 , shifting thus the breakdown point to a higher value of the residual resistivity .
all these features and experimental evidences agree well with the theory . in order to compare with other well - known cases of @xmath0-suppression by disorder
we consider the pair breaking theory for a superconductor with unconventional gap symmetry @xcite .
this physically different case of @xmath0-reduction by _ non - magnetic _ impurities in _ unconventional _
superconductors was found to be described by the famous digamma - formula of abrikosov - gorkov s conventional pair breaking theory @xcite in the presence of _ paramagnetic _ impurities .
the @xmath0-reduction may thus be expressed in the form @xmath251 where @xmath252 means the digamma function , @xmath253 is the depairing parameter , and @xmath254 the elastic scattering time . both theoretical curves , as shown in the insets of fig.[exper ] , can be chosen to coincide for weak and moderately strong disorder ( where the linear decay is rather unspecific ) .
approaching the sc - breakdown at larger disorder they differ however substantially . the agl - solution for unconventional pairing approaches @xmath255 continuously and obeys a square root dependence @xmath256 , where @xmath257 stands for the disorder - strength in the agl - case .
this square root law follows from the leading ( @xmath258 ) correction of the digamma - function @xmath259 given by the laurent series of its exponential @xmath260= \frac{\tilde{q}}{z}+\frac{z}{24\tilde{q}}+o(z^2)\ ] ] near the logarithmic branch point of @xmath252 at infinity , hence @xmath261 . in the physical context the variable @xmath262 corresponds to the critical temperature @xmath0 of eq.([digamma ] ) . by comparing the exponential of eq.([digamma ] ) one can see that @xmath263 is reached for @xmath264 .
by contrast , the present theory does not allow for a continuous breakdown of superconductivity . according to fig.3 the suppression of @xmath0 is stronger and
an abrupt breakdown occurs at @xmath265 .
the numerical data show a square root behavior however near the minimal finite @xmath0 .
the breakdown point in the curve , corresponding to our theory , seems to allow for the existence of an intermediate phase , perhaps a glassy phase below a tricritical point .
the current density is calculated according to @xmath266 , where @xmath267 . the complete expression for @xmath268 in the linear response approximation
can be obtained after averaging of @xmath269 over disorder ( [ gauss ] ) by using eq.([freeenergy ] ) for @xmath270 .
one obtains @xmath271 for the longitudinal component of the current and @xmath272 for the transverse component of the current . for simplicity
we present here only the diamagnetic contribution to the @xmath273-th ( @xmath274 ) component of the current @xmath275 where the longitudinal ( @xmath12 ) and the transverse component ( @xmath13 ) of the penetration depth are obtained as @xmath276 and @xmath277 while @xmath278 diverges at @xmath279 due to pair breaking in the sc wires , @xmath280 diverges at the global sc transition temperature @xmath82 , where the phase coherence between neighboring wires is destroyed .
the temperature and the randomness dependencies of @xmath278 and @xmath280 also strongly differ each other .
the transverse component of the penetration depth is determined by the phase - phase correlator , revealing non - linear temperature dependence and discontinuous behavior at the critical disorder strength .
nevertheless the longitudinal component of the penetration depth is given by the conventional london expression and does not depend on the disorder strength .
randomness in the josephson coupling shifts @xmath0 to lower temperatures and , therefore , the magnetic field parallel to the sc wires penetrates easier into the organic superconductor .
on the other hand , the type of disorder considered in this article does not break the cooper pairs , keeping thus the penetration of a perpendicular magnetic field into the sc wires unchanged .
in this paper we report disorder - effects on @xmath0 and on the diamagnetism of quasi-1d superconductors with random josephson - couplings .
interplay of non - magnetic disorder with quantum phase fluctuations plays a central role for the superconductor normal - metal phase transitions in this class of quasi-1d superconductors .
recent experimental data found in refs . are shown to be consistent with the present theory .
quantum criticality is controlled by two quantities , namely disorder strength and a dynamical parameter of phase fluctuations .
the present model s quantum criticality signals the existence of a quantum critical phase between sc- and normal phase .
its nature deserves further investigation .
in our study we neglect the effects of non - linear excitations , which are a subject of current interest in low dimensional systems .
note that this topic was explicitly studied by us for quasi-2d josephson coupled superconductors in ref . .
as we have shown in the previous section , the classical motion of the phase is described by a system of coupled sine - gordon type non - linear equations ( [ dynamics ] ) , which contains non - linear dynamic excitations as well as static topological defects .
the self - consistent phonon approximation allows us to calculate the phase - phase correlator between two arbitrary points @xmath281 and @xmath282 @xmath283 which can be shown to decrease at @xmath284 and @xmath285 as a power law @xmath286 , setting up a quasi - long - range order and implying the existence of a berezinskii - kosterlitz - thouless ( bkt ) topological phase transition @xcite(perhaps at @xmath2 ) in a single sc wire . the critical index @xmath287 contains both phonon and vortex contributions .
although the phonon contribution to @xmath287 can be calculated within the scha , it is not clear how the vortex contribution changes the former one . in our knowledge ,
the mechanism of excitations of the vortices with opposite fugacities and their binding in quasi-1d superconductors has not been adequately studied , and the topic needs further investigations .
this research was supported by the dfg under grant op28/7 - 1 .
e.n . thanks d. jerome for helpful communication .
r.o . is grateful for hospitality and support extended to him by the ipht at cea saclay .
in order to calculate @xmath191 for @xmath288 ( where @xmath289 are unit vectors and @xmath290 are the number of unit cells in directions of @xmath291 , correspondingly ) we rewrite eq.([s ] ) in the following form : @xmath292 where @xmath293 and @xmath294 . introducing a new variable @xmath295 and using the transformation @xmath296 one can integrate out @xmath297 and @xmath298 in eq.([s(g,0)1 ] ) . finally , @xmath299 is expressed as an integral over @xmath300 as @xmath301 \times \nonumber\\ & & \hspace{3.cm}\times \mathrm{erf } \left(\delta_{qu}^{-1}(0 ) \sqrt{u
/ 2}\right ) , \label{s(g,0)2}\end{aligned}\ ] ] where @xmath302 is the bessel function of an imaginary argument , and @xmath303 is the error function . by using the following asymptotic expressions for @xmath302 , @xmath304 , & z > >
\nu \end{array } \right.\ ] ] and for @xmath305 , @xmath306 we get for , e.g. @xmath307 at @xmath308 the following expression @xmath309 i.e. @xmath310 where @xmath311 and @xmath312 are constants of order of unity . if we take only the first terms in the expansions of @xmath313 and @xmath314 , we get @xmath315 and @xmath316 .
higher order contributions correct only these constants .
thus it is seen from eq.([s(g,0)3 ] ) that the phase - phase correlator in the transverse direction saturates at @xmath2 to its asymptotic value of @xmath317 for distances of the unit cell size @xmath129 .
the correlator @xmath318 at @xmath319 is also calculated in the same way as @xmath320 was obtained above .
using in eq.([correlatort ] ) the representation @xmath321 , the correlator @xmath318 assumes the following form @xmath322 we use the transformation @xmath323 in eq.([s(g , t)1 ] ) and carry out the integrations over @xmath324 and @xmath298 .
for the particular case of @xmath325 and @xmath326 , the expression for @xmath327 is reduced into the following form @xmath328 \times \nonumber\\ & & \times \mathrm{erf}\left(\delta_{qu}^{- 1}(t ) \sqrt{u /2}\right ) \phi \left(u , \frac{t}{\sqrt{2 } \alpha \delta_{qu}(t ) t_c^{(1)}}\right ) , \label{s(g , t)2}\end{aligned}\ ] ] where @xmath329 is given by @xmath330 here @xmath254 represents the normalized temperature ( see eq.(54 ) ) and the sum in eq.([elliptictheta ] ) is also known as the so - called _
elliptictheta_-function @xmath331 , @xcite . using the asymptotic expressions for the bessel and the error function as well as for @xmath332 in eq.([s(g , t)2 ] ) , we get the following explicit expressions for @xmath333 99 d. jrome , a. mazaud , m. ribault , and k. bechgaard , j. physique lett . *
41 * , l95 ( 1980 ) . c. bourbonnais and d. jrome ,
arxiv:0904.0617v1 [ cond-mat.str.el ] 3 apr . 2009 .
choi , p. m. chaikin , s. z. huang , p. haen , e .
m. engler , and r. l. greene , phys .
b * 25 * , 6208 ( 1982 ) . s. bouffard , m .
ribault , r. brusetti , d. jrome , and k. bechgaard , j. phys .
c * 15 * , 2951 ( 1982 ) .
b. j. powell , j. phys . : condens .
matter * 20 * , 345234 ( 2008 ) .
a. a. abrikosov , j. low temp .
phys . * 53 * , 359 ( 1983 ) .
j. shinagawa , y. kurosaki , f. zhang et al .
98 * , 147002 ( 2007 ) .
t. ishiguro , k. yamaji , and g. saito , _ organic superconductors _
( springer - verlag , heidelberg , 1998 ) .
n. joo , p. a .- senzier , c. pasquier , p. monod , d. jrome , and k. bechgaard , j. euro . phys .
b * 40 * , 43 ( 2004 ) .
n. joo , p. auban - senzier , c. r. pasquier , d. jrome , and k. bechgaard , europhys .
lett . * 72 * , 645 ( 2005 ) .
b. j. powell and r. h. mckenzie , phys .
b * 69 * , 024519 ( 2004 ) .
p. w. anderson , j. phys .
solids * 11 * , 26 ( 1959 ) .
a. a. abrikosov and l. p. gorkov , zh.eksp.teor.fiz .
* 39 * , 1781 ( 1960 ) [ sov.phys.jetp * 12 * , 1243 ( 1961 ) ] .
v. m. galitski and a. i. larkin , phys .
b * 66 * , 064526 ( 2002 ) . c. c. tsuei and j. r. kirtley , rev . mod . phys . * 72 * , 969 ( 2000 ) .
p. lee , phys.rev . lett . * 71 * , 1887 , ( 1993 ) .
p. j. hirschfeld and n. goldenfeld , phys .
b * 48 * , 4219 ( 1993 ) .
y. suzumura and h. j. schulz , phys .
b * 39 * , 11398 ( 1989 ) . e. p. nakhmedov and e. v. tahirov , j. phys .
matter * 6 * , 2245 ( 1994 ) ; ibid * 6 * 2259 ( 1994 ) .
e. p. nakhmedov and m. hner , physica c * 299 * , 301 ( 1998 ) m. ma and p. a. lee , phys .
b * 32 * , 5658 ( 1985 ) .
r. oppermann , physica a * 167 * , 301 ( 1990 ) .
a. ghosal , m. randeria , and n. trivedi , phys .
b * 65 * , 014501 ( 2001 ) .
b. spivak , a. zyuzin , and m. hruska , phys .
b * 64 * , 132502 ( 2001 ) .
v. m. galitski and a. i. larkin , phys .
lett . * 87 * , 087001 ( 2001 ) .
s. sachdev , p. werner , and m. troyer , phys .
92 * , 237003 ( 2004 ) .
b. spivak , p. oreto , and s. a. kivelson , , 214523 ( 2008 ) .
a. d. maestro , b. rosenow , m. mller , and s. sachdev , phys .
lett . * 101 * , 035701 ( 2008 ) .
f. mondaini , t. paiva , r. dos santos , and r. t. scallettar , phys .
b * 78 * , 174519 ( 2008 ) r. oppermann , z. phys .
b - cond . mat .
, 63 , 33 ( 1986 ) a. m. finkelstein , pisma zh .
* 45 * , 37 ( 1987 ) [ jetp lett . * 45 * , 46 ( 1987 ) ] .
p. w. anderson , k. a. muttalib , and t. v. ramakrishnan , phys .
b * 28 * , 117 ( 1983 ) . m. a. skvortsov and m. v. feigelman , phys . rev
. lett . * 95 * , 057002 ( 2005 ) .
k. b. efetov and a. i. larkin , zh.eksp.teor.fiz .
* 66 * , 2290 ( 1974 ) [ sov.phys.jetp * 39 * , 1129 ( 1975 ) ] .
a. firsov and g. yu .
yashin , zh.eksp.teor.fiz . *
72 * , 1450 ( 1977 ) [ sov.phys.jetp * 45 * , 761 ( 1977 ) ] .
h. j. schulz and c. bourbonnais , phys .
b * 27 * , 5856 ( 1983 ) .
v. j. emery and s. a. kivelson , nature * 374 * , 434 ( 1995 ) ; phys .
lett . * 74 * , 3253 ( 1995 ) .
e. p. nakhmedov and yu .
a. firsov , physica * c 295 * , 150 ( 1998 ) .
t. m. rice , phys .
rev.*140 * , a1889 ( 1965 ) .
the ground state of a non - interacting electron gas in an isolated @xmath86 wire or in a quasi-1d system with a nested fermi surface is a peierls dielectric .
suggestion of @xmath37- subband in the cross - section of a pure single wire surves to prevent the peierls transition in the system .
disorder in the quasi-1d system destroys also the perfect nesting , suppressing at the same time the peierls transition .
if isolated wires contain point - like impurities as well , the ground state becomes a dielectric phase due to anderson localization , and the transition becomes superconducting - insulator phase transition . in our case , anion disorder in the organic superconductors may not modify the fermi surface . according to eq.([coeff ] ) @xmath337 , this approximation is valid for @xmath338 . strictly speaking ,
this harmonic approximation is valid for small temperatures when @xmath339 , which means that few phonons are excited in the system . v. l. pokrovskii and g. v. uimin , zh .
. fiz . * 65 * , 1691 ( 1973 ) [ sov .
jetp * 38 * , 847 ( 1973 ) ] .
a. i. larkin , pisma v zh .
* 2 * , 205 ( 1965)[sov .
jetp lett.*2 * , 130 ( 1965 ) ] .
q. yuan , h .- y .
chen , h. won , s. lee , k. maki , p. thalmeier , and c. s. ting , phys .
b * 68 * , 174510 ( 2003 ) .
see , e.g. p. minnhagen , rev .
phys . * 59 * , 1001 ( 1987 ) .
mathematica handbook . | effects of non - magnetic disorder on the critical temperature @xmath0 and on diamagnetism of quasi - one - dimensional superconductors are reported .
the energy of josephson - coupling between wires is considered to be random , which is typical for dirty organic superconductors .
we show that this randomness destroys phase coherence between wires and that @xmath0 vanishes discontinuously at a critical disorder - strength .
the parallel and transverse components of the penetration - depth are evaluated .
they diverge at different critical temperatures @xmath1 and @xmath0 , which correspond to pair - breaking and phase - coherence breaking respectively .
the interplay between disorder and quantum phase fluctuations is shown to result in quantum critical behavior at @xmath2 , which manifests itself as a superconducting - normal metal phase transition of first - order at a critical disorder strength . |
to date , the large magellanic cloud ( lmc ) is the only normal galaxy other than the milky way that has been seen in high energy , diffuse emission ( sreekumar et al .
most normal galaxies are simply too distant , and their cosmic ray ( cr ) densities are too small to have sufficient emissivities .
starburst galaxies , however , have many supernovae due to their unusually high rates of massive star formation .
the supernovae produce copious cr electrons seen as bright , extended synchrotron regions ( e.g. , seaquist & odegard 1991 ; carilli et al .
in addition , starburst nuclei typically contain high masses of dense molecular gas ( young et al .
1989 ; solomon , downes & radford 1992 ; helfer & blitz 1993 ; paglione et al .
1997 ) . since crs are the likely cause of the high temperatures derived deep in the cores of these clouds ( suchkov , allen & heckman 1993 ) , production , through interactions between gas and crs , should be relatively high in starburst galaxies .
the challenge has been to accumulate enough exposure time to detect the emission . as the archetypal starburst galaxies ( rieke et al .
1980 ) , the nearby galaxies ngc 253 and m82 have been the primary targets of searches .
both have high supernova rates of @xmath3 yr@xmath4 derived from high resolution radio continuum images ( van buren & greenhouse 1994 ; ulvestad & antonucci 1997 ) .
co maps indicate molecular gas column densities of @xmath5 within their inner kiloparsec ( paglione et al .
1999 ; carlstrom 1989 ) , roughly 10 times that of the milky way ( e.g. sanders , solomon , & scoville 1984 ) . studies of their dense molecular gas indicate high average densities of @xmath6 in m82 to @xmath7 in ngc 253 ( paglione et al .
therefore , ngc 253 and m82 are expected to be the brightest starburst galaxies in .
however , previous egret studies did not detect these objects at high energies ( sreekumar et al .
1994 ; paglione et al .
1996 ) .
more cgro data have become available in the meantime which should improve sensitivity . by some accounts ( pohl 1994 )
, even these starburst galaxies may not be luminous enough in to be detected individually by egret with the exposure currently available .
however , a combination of all observations obtained for a _ collection _ of starburst galaxies may add up to a detectable signal . in this paper
we present such a search from a survey of 9 nearby starburst galaxies .
in addition we analyze observations of the nearby normal galaxy m31 .
the spark - chamber on board cgro operates in the @xmath8 mev to @xmath8 gev energy range .
the instrument is capable of producing circular @xmath9-ray images with typical radii of 30 .
the location accuracy for point sources is about 05 .
a detailed description of the instrument and its performance is given by hughes et al.(1980 ) and kanbach et al .
the calibration properties are described by thompson et al .
( 1993 ) .
sky maps are obtained by applying a maximum - likelihood method ( mattox et al .
1996 ) , which provides flux estimates and statistical significances of sources identified in the data .
the instrumental background of is small compared to the measured celestial emission .
the celestial background in the data is dominated by an isotropic component , which is most likely diffuse extragalactic emission ( further discussed in sreekumar et al .
1998 ) . a flat background model which allows for this isotropic emission
is created for each observation and energy range . for observations near the galactic plane
, we additionally take into account galactic diffuse emission by creating models for the hi and co distributions , and for the inverse - compton ( ic ) radiation .
these models are made by convolving maps of the gas and photon distributions with an point - spread function for each energy interval .
we apply hi maps based on various 21-cm surveys ( combined into an all - sky map ; see bloemen et al .
1986 ) , and the co map of dame et al .
the ic distribution is derived from a cosmic ray propagation model by strong & youssefi ( 1995 ) .
we apply the maximum - likelihood analysis software originally developed to analyze data from the comptel @xmath9-ray telescope aboard cgro .
this software has been augmented to analyze archival data ( summarized in blom 1997 ) , and has been verified to give flux values and significances consistent with published results of the team .
we use individual archival viewing periods rather than the all - sky mosaic supplied by the cgro science support center , because they cover a broader time span ( to make our search as sensitive as possible we need the maximum exposure available ) .
the fluxes we derive for several blazar sources near the galactic plane are consistent with the published results , which indicates that our diffuse models do not introduce significant deviations . for example
, the blazar 2eg j1730 - 130 at @xmath10 was marginally detected by egret above 100 mev during a two - week pointing in february 1992 at a flux level of @xmath11 ph ( thompson et al .
1995 ) , or @xmath12 ph based on an improved analysis ( mukherjee et al .
we find @xmath13 ph , which is in good agreement with the two flux values published by the egret team . in order to combine different observations for a particular source , we select data from all pointings taken within 30 of the source location and coadd the count distributions and exposure maps separately .
the coadding procedure is straightforward only when the viewing periods have matching coordinate systems .
the egret archives usually provide data in galactic coordinates when observations are performed near the galactic plane , otherwise viewing periods are given in equatorial coordinates .
for most starburst sources studied in this paper , observations with consistent coordinates are available in the archives . however , in a few cases we are forced to combine data in galactic coordinates with data in equatorial coordinates .
since we are only interested in the signal at one pixel ( the source location ) , we basically ignore the coordinate system .
in other words , we shift the origins of the data sets to the source location and subsequently coadd the files assuming flat coordinates .
the same procedure is used to coadd the signals from a _ collection _ of sources .
note that we have verified that the mosaics obtained from both coadding procedures are reliable in the sense that analyses of combined blazar observations yield flux values that agree with published results .
when we fit a point - source model to combined data at sky locations one or two pixels away from the true source location ( 0.5 to 1 off ) , we obtain flux values that are lower than , but consistent with the optimal value .
this result shows that our coadding procedure is not sensitive to small coordinate / binning mismatches between viewing periods .
the observation program of cgro consists of four phases .
phases i to iii each consist of @xmath14 yr of observations . phase iv is subdivided into annual cycles , starting with cycle iv .
data from phase i up to phase iv / cycle v , covering the period 19911996 , are used in this work .
table [ tab.sou ] shows the relevant egret observations for each galaxy in our survey .
specific observing dates of the cgro phases and viewing periods are given elsewhere ( e.g. , gehrels et al . 1994 ) .
the survey consists of ten starburst galaxies selected by distance ( @xmath15 10 mpc ) , iras far infrared ( fir ) luminosity ( @xmath16 , young et al .
1989 ) , and angular distance from the galactic plane ( @xmath17 ) .
we omit one candidate , ngc 891 , because of its proximity to the egret source 2eg j0220 + 4228 ( thompson et al . 1995 ; mattox et al .
the large spiral galaxy m31 is analyzed separately as well .
though its fir luminosity ( and therefore the supernova rate ) is an order of magnitude lower than a starburst galaxy s , m31 may be close enough to make up for its lower emissivity .
all observations listed in table [ tab.sou ] for ngc 253 and m82 are combined separately .
for each source , maximum - likelihood ratio maps are created in six standard egret energy ranges ( 70100 mev , 100150 mev , 150300 mev , 300500 mev , 0.51 gev , and 12 gev ) , and in one broad energy range covering all photon energies @xmath18 mev .
flux values are derived by simultaneously fitting point - source models at the locations of the starburst galaxy and all respective neighboring egret catalog sources within @xmath19 on the sky ( thompson et al .
1995 , 1996 ) .
none of the likelihood maps show evidence for emission from ngc 253 or m82 .
upper limits ( @xmath20 ) obtained from model fitting are shown in figure [ fig.gammas ] .
note that no upper limit is derived for m82 at 12 gev due to limited statistics . for ngc 253
we find a @xmath20 upper limit to the emission above 100 mev of @xmath21 ph @xmath22 s@xmath4 . for m82
this limit is @xmath23 ph @xmath22 s@xmath4 .
note that both limits are well below previous estimates ( sreekumar et al .
1994 ; paglione et al .
1996 ) , which further constrains emission models ( [ model ] ) .
we apply the same analysis technique for m31 at @xmath24 mev only .
no evidence for m31 is found in this energy range .
we find a 2@xmath0 upper limit for the flux above 100 mev of 1.6 ph .
this value is below the estimate of zel & berkhuijsen ( 1987 ) . in order to create a `` supermosaic '' for all starburst galaxies listed in table [ tab.sou ] ,
the observations for each source are first combined separately .
the resulting submosaics are then coadded such that the locations of all starburst galaxies match in one pixel ( with arbitrary coordinates ) .
in addition , the event distributions of the submosaics are scaled to ensure that each starburst galaxy contributes equally in exposure to the supermosaic .
all background models are treated similarly .
we create maximum - likelihood ratio images for the supermosaic in six standard egret energy ranges between 70 mev and 2 gev , and for all photon energies @xmath24 mev .
these images show significant excesses that can be identified with the source signals of strong egret blazars which remain visible in the supermosaic even after adding noise from observations that do not contain these sources .
however , no evidence for emission at the shared location of the starburst galaxies is found .
upper limits derived from simultaneous model fitting taking the egret blazars into account are shown in the right panel of figure [ fig.gammas ] .
the @xmath20 upper limit we derive for the _ average _ emission above 100 mev from starburst galaxies is @xmath25 ph @xmath22 s@xmath4 .
to calculate the flux from a starburst galaxy , we model the cr production and radiation mechanisms for ngc 253 and m82 .
a full description of the model is given in paglione et al .
( 1996 ) . in short ,
we assume a power law form for the initial cr injection spectrum @xmath26 , and allow it to evolve to a steady state . the ambient density ( with a main contribution from in galactic nuclei ) , magnetic field and photon field determine the energy loss rates due to bremsstrahlung , synchrotron and ic radiation , as well as secondary particle production . the model yields steady state primary and secondary electron , and proton energy distributions , which are used to calculate synchrotron and spectra .
the normalization of the proton injection spectrum is estimated by comparing the modeled and observed synchrotron radio spectra .
this normalization corresponds to the total power per unit time transferred by supernovae into cosmic rays within a given volume , @xmath27 here @xmath28 is the efficiency of energy transfer from a supernova to crs , @xmath29 ergs is the supernova energy , @xmath30 is the supernova rate , and @xmath31 is the volume . solving for the normalization @xmath32 , we find ( for a minimum energy @xmath33 and a power law slope of 2.2 ) , @xmath34 a minimum cr energy , @xmath35 ( bell 1978 ) , is required for acceleration in a shock front .
it is roughly a few mev for a shock velocity of @xmath36 . to generate fluxes ,
the model results are integrated over volume and divided by 4@xmath37 , where @xmath38 is the galaxy distance . for ngc 253 ,
we choose a disk - shaped starburst region 70 pc thick , with a radius of 325 pc ( paglione et al .
1999 ) , and @xmath39 yr@xmath4 ( ulvestad & antonucci 1997 ) . for m82 , @xmath40 yr@xmath4 ( van buren & greenhouse 1994 ) , and
the disk is 50 pc thick with a radius of 355 pc ( shen & lo 1995 ) .
paglione et al .
( 1996 ) tested for emission outside the starburst region and found it to be negligible , so we ignore any interactions outside the nucleus .
the calculated synchrotron spectra resulting from the model cr electron distributions match the shape of the observed radio spectra from the nuclei of ngc 253 and m82 ( carilli 1996 ; carlstrom & kronberg 1991 ) given a photon field of 200 ev and a diffusion time scale of 10 myr ( cf .
paglione et al .
the average gas densities for ngc 253 and m82 are 300 and 100 , respectively .
emission measures of @xmath41 and 10@xmath42 @xmath43 pc are required to match the free - free component and thermal absorption ( low frequency turn - over ) of the spectra from ngc 253 and m82 , respectively ( cf .
, carilli 1996 ; carlstrom & kronberg 1991 ) . on the whole ,
the density and photon field are well constrained by the radio spectra .
the density range is limited by the slope of the radio spectrum .
higher densities increase the cr losses due to collisions with matter , thus reducing the number of low energy crs , resulting in a flatter synchrotron spectrum .
( note that thermal absorption only affects the spectrum at low frequencies and is easily distinguished from the inherent slope of the underlying emission . )
the volume - averaged densities that best match the radio spectra agree well with those found from molecular line studies ( paglione et al .
the photon field is determined by the fir luminosity and , to a much lesser degree , by the synchrotron slope .
a steeper slope indicates a higher photon field since more high energy electrons are lost due to ic radiation .
the absolute level of the model synchrotron spectrum is determined by the magnetic field @xmath44 , the ratio of proton and electron densities @xmath45 , and @xmath28 . though the magnetic field , derived from radio data , is most likely @xmath46 g in ngc 253 and m82 ( carilli 1996 ; seaquist & odegard 1991 ) , @xmath45 and @xmath28 are unknown
. the @xmath45 ratio can be anywhere between 20 and 2000 , though values of 50 to a few hundred appear most likely for the milky way ( bell 1978 ) .
values for the efficiency of energy transfer between 10 and 50% have been found from models of the evolution of supernova blast waves ( markiewicz , drury & vlk 1990 ) .
fortunately , the flux depends only very weakly on @xmath44 and @xmath45 .
therefore , @xmath28 is the most important parameter needed to predict the flux , and it is determined from the radio spectrum normalization .
five models ( a e , table [ tab.models ] ) are shown with the egret upper limits in figure [ fig.gammas ] .
the spectra are summations of ic , bremsstrahlung and neutral pion decay radiation . near 100 mev ,
the emission is dominated by neutral pion decays as seen by the so - called `` pion bump . ''
though all the models lie below the m82 limits , model a predicts a flux above 100 mev for ngc 253 higher than measured . from their integrated radio spectra and molecular gas properties ,
the other galaxies in the survey are relatively similar to ngc 253 and m82 .
we assume that the major difference in the normalization between starburst galaxies is the supernova rate and distance ( equation [ eq.k ] ) .
we estimate the supernova rate from the fir luminosity ( van buren & greenhouse 1994 ) . for ngc 4945 , a starburst with very high gas densities ( jackson et al .
1995 ) , we use the model fit to ngc 253 . for the other starburst galaxies , we use the model for m82 .
the expected average spectra from models a e are shown in figure [ fig.gammas ] with the egret upper limits .
unlike the models for ngc 253 , which are constrained by the egret points near the pion bump , these results are limited by the data near 1 gev .
though many of the model fluxes lie below the egret limits , all but one or two models may be eliminated . the low magnetic field in model a contradicts minimum energy estimates from radio data ( beck et al .
1994 ; rieke et al .
models a and b both have very high transfer efficiencies , and their spectra are above or very near near the 2@xmath0 limits in a few energy bands for ngc 253 and the starburst galaxies . with model b eliminated , we conclude that @xmath47 , as in the milky way . note that model b may still be valid for m82 , though we have no reason to believe that the cr population would be so different in this particular galaxy .
the magnetic field used for model c agrees with previous calculations , and the @xmath45 ratio of 100 matches that observed for the milky way near 1 gev .
the integrated fluxes and spectra all fall below the egret limits . after reviewing the calculation of the minimum energy magnetic field
, we find model d to be a poor solution . from the equation for magnetic field strength in beck et al .
( 1994 ) , we find a gas volume filling factor in ngc 253 of @xmath48 for model d. this result implies ionized cloud densities of roughly @xmath49 . since these densities approach those found for the dense molecular medium in ngc 253 ( paglione et al .
1999 ) , model d may not be appropriate due to its low volume filling factor .
assuming similar magnetic fields in m82 and the starburst galaxies , model d is most likely not valid for them either .
another argument against model d comes from comparing its predictions with the expected milky way flux .
the luminosity of the milky way measured by egret above 100 mev is roughly 2 ph(s .
hunter , private communication ) , and its fir luminosity is 7(cox & mezger 1989 ) .
when we scale the milky way luminosity to the average starburst fir luminosity ( which is roughly 4 times higher ) and distance ( 7.4 mpc ) , we would observe a flux of @xmath50 ph .
this is nearly the flux predicted for the survey sources with model d. since the emissivity should be higher in starburst galaxies due to their larger average gas densities , this simple scaling most likely underestimates the true flux .
we therefore consider model d at best a lower limit prediction .
model e is eliminated since the predicted fluxes all fall below the scaled estimate from the milky way .
further , the flux of the lmc ( sreekumar et al .
1992 ) , when similarly scaled , is consistent with , or slightly above , the fluxes from model e. the emissivity of the lmc is much lower than in starburst galaxies . at a distance of 50 kpc , even the flux of the milky way is already 35 times higher than in the lmc .
therefore , the fluxes predicted by model e are clearly too low .
note that this result does not strictly eliminate @xmath51 with this ratio we can obtain fluxes that are similar to model d given @xmath52 and @xmath53 g .
based on our new analysis and egret limits , we expect the flux estimates from model c to be the most realistic ( with model d as a lower limit ) .
these models imply @xmath54 , similar to the galactic value for cr energies @xmath55 gev .
the energy from supernovae is also efficiently transferred to crs ( @xmath5650% ) .
the survey sources have magnetic field strengths @xmath57 g .
the egret limits also constrain the relative supernova ( or massive star formation ) rate in the starburst galaxies . at the average distance of our survey sources
, the flux above 100 mev from the milky way would be 3 ph .
compared with the observed starburst upper limit , we find @xmath58 .
this limit is conservative since not even a hint of starburst emission is found in the egret maps .
the integrated fluxes from models c and d indicate @xmath59 and the difference in fir luminosities implies @xmath60 .
given the uncertainties in @xmath28 , we estimate that the star formation rate in starburst galaxies is only about an order of magnitude higher than in the milky way .
studies at other wavelengths also indicate star formation rates elevated by this magnitude ( e.g. , kennicutt 1998 ) .
it is interesting to note that at the distance of m31 , the flux of the milky way above 100 mev would be 4 ph .
this flux is over twice the egret limit for m31 .
therefore , the galactic emissivity must be higher than in m31
. three main differences between our galaxy and m31 can explain this result .
first , the supernova rate in the milky way is roughly 7 times higher than in m31 , as can be estimated from the fir luminosity .
this difference is probably even more pronounced because much of the fir emission in m31 originates from population ii stars in the nucleus rather than from massive star forming regions ( dame et al .
1993 ) . therefore ,
if crs are in fact accelerated in the shock fronts of supernovae , the cr density in the milky way , and thus its emissivity , should be higher .
second , the mass of and in the milky way is at least 6 times that of m31 ( e.g. , dame et al .
1993 ; clemens , sanders & scoville 1988 ) .
third , unlike the starburst galaxies and the milky way , m31 lacks a massive concentration of dense molecular gas in its nucleus ( though some unexplained radio emission is seen there ) .
this implies a lower secondary particle production in m31 ( including neutral pions ) which means a lower diffuse flux .
however , the emissivity does not depend very strongly on density , so the difference in supernova rates is the most important factor determining the high energy emission from galaxies . we conclude that our null detection of hard emission from m31 , compared with the predicted milky way flux at the same distance , supports the theory of the supernova origin of crs . throughout the paper we have neglected point - source contributions to the high - energy emission from starburst galaxies , m31 , and the milky way . in principle , a large population of relatively weak sources
may add significantly to the total high - energy emission of a galaxy .
studies on the nature of the many unidentified egret sources suggest that supernova remnants ( without an embedded pulsar ) , x - ray binaries and flare stars are all possible , but not yet well - established sources of high - energy ( mukherjee , grenier & thompson 1997 ) .
pulsars are the only firmly identified small - scale objects that are known to emit up to gev energies .
therefore , we calculate the emission from a galactic ensemble of pulsars to estimate the possible contribution of such a source population to the overall luminosity in a galaxy .
egret has detected 7 pulsars above @xmath61 mev in the galaxy which show an increasing efficiency in converting rotational energy of the spinning neutron star into versus pulsar age .
a second obvious trend is an increasing spectral hardening in the regime versus age ( fichtel & trombka 1997 ) .
their typical pulsed spectra can be described by power - laws which cut off or break at @xmath62@xmath63 gev ( thompson et al .
all _ radio _ pulsars in the ( fairly complete ) princeton pulsar catalog ( taylor , manchester & lyne 1993 ) may generally have similar emission properties .
when we assume that the radio and emission cones of pulsars have the same aperture , we can easily estimate an integrated pulsar contribution to the high - energy emission of the galaxy above 100 mev .
we find a maximum contribution of @xmath64 when we adopt a long - lived pulsar activity ( @xmath65 yr ) and an average spectral cut - off at low energies ( @xmath66 gev ) .
this result indicates that pulsars add an insignificant fraction to the total luminosity of the galaxy above 100 mev .
this is in agreement with the findings of pohl et al .
( 1997 ) , who conclude that only for energies @xmath67 gev we may find a significant contribution ( up to @xmath68 ) in selected sky regions .
the flux values listed in the egret catalogs imply that the integrated emission from unidentified sources is comparable to that from pulsars . if all the egret unidentifieds are galactic sources of some type other than pulsars ( which is in fact unlikely ) , then we have a maximum total point - source contribution of @xmath69 to the overall emission @xmath24 mev in the galaxy . since we have no reason to expect a more dominant contribution from pulsars ( and other point sources ) in starburst galaxies or m31 , we conclude that it is justified to neglect point - source contributions in our model predictions .
our combined egret observations are the most sensitive measurements of starburst galaxies ever performed at hard energies
. however , no mev or gev emission from starburst galaxies is identified .
nevertheless , we have shown that the derived egret upper limits are useful for constraining models of cr and production in galaxies .
the modeling implies large average gas densities , magnetic fields , and photon fields .
we find @xmath70 , similar to the milky way value , and a high efficiency of energy transfer from supernovae to crs ( 2050% ) .
these values agree well with radio observations . according to our model predictions ,
an instrument with an order of magnitude better sensitivity than egret should be able to detect ngc 253 and m82 in a one - year all - sky survey . the gamma - ray large area space telescope ( glast ; e.g. , michelson 1996 ) , with a projected point - source sensitivity in one year of 5 ph , may have this capability .
the glast pair conversion telescope is being designed to image the 10 mev 300 gev sky .
its launch is expected in the first decade of the next century .
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 and space administration .
j. j. b. acknowledges computational support from the space research organization netherlands ( sron ) in utrecht , the netherlands .
we thank s. bloom and s. hunter for helpful discussions .
lccccl m31 & 0.67 & 1211 & @xmath71216 & 0.1@xmath72 & 26.0 , 28.0 , 34.0 , 37.0 , 211.0 , & & & & & 325.0 , 401.0 , 425.0 , 530.0 & & & & & * total exposure : * @xmath73 @xmath74 s ic 342 & 1.8 & 138.2 & 10.6 & 0.2 & 15.0 , 18.0 , 31.0 , 211.0 , 216.0 , & & & & & 319.0 , 319.5 , 325.0 , 401.0 , 411.0 , & & & & & 411.5 , 518.5 , 530.0 & & & & & * total exposure : * @xmath75 @xmath74 s m82 & 3.25 & 141.4 & 40.6 & 3.2 & 18.0 , 22.0 , 216.0 , 227.0 , 228.0 , 319.0 , & & & & & 319.5 , 411.0 , 411.5 , 418.0 , 518.5 & & & & & * total exposure : * @xmath76 @xmath74 s ngc 253 & 3.4 & 97.4 & @xmath7188.0 & 2.8 & 9.0 , 13.5 , 404.0 , 428.0 & & & & & * total exposure : * @xmath73 @xmath74 s m83 & 5.4 & 314.6 & 32.0 & 2.0 & 12.0 , 23.0 , 32.0 , 207.0 , 208.0 , & & & & & 215.0 , 217.0 , 316.0 , 405.0 , 405.5 , & & & & & 408.0 , 424.0 , 511.5 & & & & & * total exposure : * @xmath77 @xmath74 s ngc 2903 & 7.0 & 208.7 & 44.5 & 0.8 & 40.0 , 322.0 , 326.0 , 403.5 & & & & & * total exposure : * @xmath73 @xmath74 s ngc 4945 & 7.0 & 305.3 & 13.3 & 8.5@xmath72 & 12.0 , 14.0 , 23.0 , 27.0 , 32.0 , & & & & & 207.0 , 208.0 , 215.0 , 217.0 , 230.5 , & & & & & 303.0 , 314.0 , 315.0 , 316.0 , 402.0 , & & & & & 402.5 , 424.0 , 522.0 , 531.0 & & & & & * total exposure : * @xmath78 @xmath74 s m51 & 9.7 & 104.8 & 68.6 & 3.1 & 4.0 , 22.0 , 218.0 , 222.0 , 418.0 , 515.0 & & & & & * total exposure : * @xmath73 @xmath74 s ngc 6946 & 11.0 & 95.7 & 11.7 & 4.8 & 2.0 , 34.0 , 203.0 , 212.0 , 302.0 , & & & & & 303.2 , 303.7 , 318.1 , 401.0 , 530.0 & & & & & * total exposure : * @xmath79 @xmath74 s ngc 3628 & 11.5 & 169.4 & 13.9 & 1.9 & 3.0 , 4.0 , 11.0 , 204.0 , 205.0 , 206.0 , & & & & & 222.0 , 304.0 , 305.0 , 306.0 , 307.0 , & & & & & 308.0 , 308.6 , 311.0 , 311.6 , 312.0 , & & & & & 313.0 , 322.0 , 326.0 , 511.0 , 515.0 & & & & & * total exposure : * @xmath80 @xmath74 s lcccccc a & 25 & 50 & 0.90 & 36.5 & 37.5 & 14.4 b & 50 & 400 & 0.73 & 24.1 & 27.6 & 10.0 c & 50 & 100 & 0.43 & 15.6 & 16.7 & 6.3 d & 100 & 100 & 0.15 & 5.3 & 6.1 & 2.2 e & 100 & 50 & 0.05 & 2.0 & 2.2 & 0.8 | we present a search for high energy gamma - ray emission from 9 nearby starburst galaxies and m31 with the energetic gamma - ray experiment telescope ( egret ) aboard the compton gamma - ray observatory ( cgro ) .
though the diffuse gamma - ray emission from starburst galaxies was suspected to be detectable , we find no emission from ngc 253 , m82 nor from the average of all 9 galaxies .
the 2@xmath0 upper limit for the egret flux above 100 mev for the averaged survey observations is 1.8 ph . from a model of the expected radio and emission ,
we find that the magnetic field in the nuclei of these galaxies is @xmath1 g , and the ratio of proton and electron densities is @xmath2 .
the egret limits indicate that the rate of massive star formation in the survey galaxies is only about an order of magnitude higher than in the milky way .
the upper limit to the flux above 100 mev for m31 is 1.6 ph . at the distance of m31 , the milky way flux would be over twice this value , indicating higher emissivities in our galaxy .
therefore , since the supernova rate of the milky way is higher than in m31 , our null detection of m31 supports the theory of the supernova origin of cosmic rays in galaxies .
# 1 # 1 # 1#2 # 1 # 1#2 # 1 # 1 # 1 # 1 # 1 # 1 |
gauge theories were led to a major position in the foundations of physics as the main tool to describe the standard model of elementary particles . a great variety of work in this area has been done , and important achievements have been reached by the study of non - abelian symmetry groups such as @xmath0 or @xmath1
. however , deeper analyses are still necessary , specially in non - perturbative cases , in order to understand subtle aspects and to discover new properties of these theories .
hence , an important tool to describe these systems is the semiclassical limit , with which it is possible to obtain valuable insights about their quantum behaviour . and , in this context , an approach that is largely used is to analyze the orbits of a single particle , which is the main objective of this article .
we start proposing an expression for the quadriforce that describes the interaction between a particle with isospin and fields from gauge theories , which usually have scalar and vector components .
although such expression is well established in the case of vector fields ( the so called wong s expressions @xcite ) , it is still a matter of investigation for scalar fields ( @xcite , @xcite ) . to obtain an expression that takes into account both contributions ,
we make use of wong s equation combined with a gauge generalization of the quadriforce presented in our previous work @xcite , which referred to scalar fields only . with this formalism
, we apply the resultant equations to two cases of physical significance : a coloured particle interacting with a bps monopole , which behaves asymptotically as a magnetic monopole , and with a julia - zee dyon , which behaves asymptotically as a magnetic plus electric monopole .
we observe that such systems present no dynamical equilibrium points and have no apparent general analytical solutions .
hence , we particularize to simpler cases in which the general properties of the motion of the particle can be well established : radial and asymptotic motions .
as far as the radial motion is concerned , we find that there are no bounded orbits for the particle : it always escapes in a direction that is related to the sign of the radial component of its isospin . analyzing the asymptotic motion , in the limits of low ( newtonian ) and high ( ultra - relativistic ) velocities
, we observe the existence of bounded orbits under certain conditions . the remarkable property , however , is that all of the asymptotic trajectories are constrained to a conical surface in both limits . and this is exactly the case of classical electromagnetism , in which the motion of a point charge interacting with a magnetic monopole is restricted to the surface of a cone .
this article is divided as follows : in section [ sec : bps ] we derive the equations of motion for the particle in the presence of the bps monopole and analyze it globally ; in section [ sec : radial ] , we specify them to radial motion and , in section [ sec : asympt ] , to asymptotic motion .
the case in which the monopole involved is the julia - zee dyon is presented in section [ sec : dyon ] and conclusions are described in section [ sec : concl ] .
firstly , we derive the general equations that describe the interaction of a particle with isospin and gauge fields
. two sets of equations are requested : the one referring to the quadriforce and the one to the isospin s precession .
as we have previously explained , we propose an expression for the quadriforce that is a combination of wong s equation and a generalization of the equation proposed in our last article .
this was originally conceived to take into account scalar fields only : @xmath2 in the above expression , @xmath3 are the coordinates of the particle , @xmath4 is its proper time , @xmath5 are the components of its isospin , @xmath6 is the coupling constant , @xmath7 is the minkowskian metric and @xmath8 are the scalar fields .
it is straightforward to generalize ( [ original ] ) to gauge fields : we only have to replace the ordinary derivatives by covariant ones : @xmath9 to include the self - interaction of the gauge fields as well , it is necessary to introduce wong s expression , which leads us to : @xmath10 where @xmath11 are the components of the strength tensor of the fields . the isospin s precession equation is obtained just in the same way : we make use of the expression proposed in our last work generalized to gauge fields ( which recovers wong s expression in the case of vector fields only ) : @xmath12 therefore , equations ( [ quadriforce ] ) and ( [ isospin ] ) are the equations of motion of a coloured particle interacting with gauge fields .
the bps monopole is described by the fields ( @xcite , @xcite ) : @xmath13 where : @xmath14 and @xmath15 is a constant .
hence , by applying them to the equations ( [ quadriforce ] ) and ( [ isospin ] ) , we get the following system : @xmath16 + \frac{hk}{r^2 } \left ( \vec{i } \times \vec{v } \right ) + \frac{rh'-h ( 1+k)}{r^2 } \left ( \vec{r } \cdot \vec{v } \right ) \left ( \vec{i } \times \vec{r } \right ) \nonumber\end{aligned}\ ] ] we can write it as a dynamical system and look for equilibrium points to study its stability . imposing the equilibrium condition and using a base of co - moving coordinates , @xmath17 it is straightforward to conclude that there are no equilibrium points .
therefore , we particularize ( [ system ] ) to more restricted conditions and study the general properties of the resultant orbits .
if the particle is initially moving in the radial direction , @xmath18 , and its isospin is also in the same direction , @xmath19 , we can use equations ( [ system ] ) to show that the particle will remain in the radial direction according to : @xmath20 where the function @xmath21 is defined as : @xmath22\ ] ] to study this unidimensional motion , we can firstly analyze it in the newtonian limit , @xmath23 .
it is clear that , in this case , the function @xmath21 is nothing but the potential energy of the particle due to its interaction with the monopole ; moreover , this function has the following properties : @xmath24 such characteristics of the potential energy can also be seen in figure 1 , in which the graphic of the function is shown for the case @xmath25 .
hence , we conclude that there are no bounded orbits in this limit , and the particle will always escape . of course , in order for the particle not to leave the region of validity of the newtonian limit , it is necessary to guarantee that the constants involved satisfy the condition @xmath26 .
now , we can analyze equations ( [ radial ] ) in the relativistic case .
although an analytical solution is not available , we can infer the main properties of the resultant motion by seeking a constant of motion .
taking @xmath27 , the total relativistic kinetic energy of the body , we obtain from ( [ radial ] ) the following constant of motion : @xmath28 this can be interpreted as the total relativistic energy of the particle , since that , in the newtonian limit , the first term reduces to the well known expression of the classical kinetic energy . analyzing this constant ,
it is clear that there is no more than one return point ( i.e. , points in which @xmath29 ) in the system , implying that no bounded orbits are allowed , and again the particle always escapes .
the equation describing such motion is given implicitly by : @xmath30 where : @xmath31 an interesting feature of equations ( [ system ] ) is their behaviour when @xmath32 , since , in this limit , the bps solution behaves as a magnetic monopole . applying this condition ,
we are led to the new system : @xmath33 in which , again , @xmath34 is the radial component of the particle s isospin .
these equations can indeed be simplified by changing the coordinate system to that of ( [ base ] ) ; hence , writing the isospin vector as : @xmath35 we obtain that the second equation of ( [ asympt ] ) has the following solution : @xmath36 where @xmath37 . thus , the motion of the particle in the external ( minkowskian ) space is not coupled to the motion in the internal ( isospin ) space , where its isospin precessionates according to the previous expressions .
so , we have to solve the first equation of ( [ asympt ] ) to completely describe the particle s movement .
however , we seem to have neither an analytical solution nor enough constants of motion for the equation . hence
, we make further approximations and study the system in two very distinct limits : the newtonian limit ( @xmath38 ) and the ultra - relativistic limit ( @xmath39 ) . in the newtonian limit
, the equation is reduced to : @xmath40 we observe that ( [ assintnewton ] ) is very similar to the lorentz equation of electromagnetism corresponding to a particle with charge @xmath34 interacting with an electric field generated by a unit charge and a magnetic field generated by a unit magnetic monopole .
the first term is due to the scalar fields and , the second , due to the vector fields .
equation ( [ assintnewton ] ) has three constants of motion that can be directly obtained : @xmath41 it is clear that the first one is the total energy of the particle , the second is its total angular momentum ( see , for instance @xcite ) and the third , the module of its orbital angular momentum . to determine the equations of motion , we choose spherical coordinates and make the @xmath42 axis coincide with @xmath43 .
therefore , it is straightforward that the particle is restricted to move on the surface of a cone whose axis is in the same direction of @xmath43 and whose half - angle of opening is given by : @xmath44 in addition , combining the two other remaining constants , we obtain : @xmath45 it is now possible to analyze the motion in the radial direction by this single constant .
the shape of the effective potential is shown in figure 2 , considering the two possible signs of @xmath34 .
we can conclude that , if @xmath46 , the particle s energy has to be positive and there is only one return point in : @xmath47 it is important to notice that this `` return point '' is actually a point in which @xmath48 vanishes , and not the velocity @xmath49 . therefore , the particle will also escape .
moreover , it is necessary that @xmath50 and @xmath51 for the return point to be in the asymptotic region and for the particle to remain in the newtonian limit , respectively . however , in the case in which @xmath52 , the particle is allowed to have positive or negative energy .
if it is positive , there is again only one return point given by : @xmath53 meaning that there are no bounded orbits .
but if the energy takes negative values , there are two return points , given by : @xmath54 which means that the movement is bounded . in addition
, there is one equilibrium point , in which the particle just precessionates along the intersection between the cone and the sphere of radius given by : @xmath55 of course , some conditions regarding all these constants have to be satisfied in order to guarantee that these return and equilibrium points are in the asymptotic region and that the particle remains with low velocities .
these are : @xmath56 , @xmath57 , @xmath58 and @xmath59 .
the analytical expression of the radial coordinate , @xmath60 , can be obtained directly by solving the differential equation ( [ vefetivo ] ) .
we note that , if @xmath61 , the solution is : @xmath62 ^{r(t)}_{r_0}= \pm t \end{aligned}\ ] ] while , for @xmath63 , it is given by : @xmath64 ^{r(t)}_{r_0}= \pm t\end{aligned}\ ] ] the sign in the equations has to be chosen according to the initial radial velocity of the particle and changes every time it crosses one of the return points presented .
to complete the description of the movement , we have to determine the expression for the azimuthal angle , @xmath65 .
writing the total velocity in spherical coordinates , @xmath66 , and using the constants of motion , we arrive at the implicit equation : @xmath67 in the case where the particle precessionates along the intersection between the cone and the sphere of radius ( [ pontoequil ] ) , this last equation can be explicitly solved , and the result is given by : @xmath68 now that we have analyzed the motion in the newtonian limit , we can move to the opposite limit , the ultra - relativistic motion . in this case , the equation of motion is reduced to : @xmath69 it is clear that the velocity of the particle is a constant of motion , implying that once the body is in the ultra - relativistic limit , it will remain there .
moreover , two other constants can be readily obtained : @xmath70 they are , respectively , the total relativistic angular momentum and the module of the orbital relativistic angular momentum .
hence , the orbits of the particle are again confined to a conical surface whose axis coincides with @xmath71 and whose half - angle of opening is : @xmath72 by using the fact that the module of the orbital angular momentum and the velocity are constants , it is possible to get a differential equation for @xmath60 whose solution is given explicitly by : @xmath73 where @xmath74 is the initial angle between the velocity and the position of the particle and the sign of the equation has to be chosen according to the initial sign of the radial velocity of the particle .
we can see that , even if the particle is initially going in the direction of the origin , it will never reach this point , but will stop at @xmath75 and return , escaping to infinity .
therefore , in this limit of high velocities , there is no bounded motion . to complete the description of the particle s movement ,
it is straightforward to obtain the azimuthal coordinate explicitly by using the constants of motion and the equation for @xmath60 : @xmath76\ ] ]
now , we can apply the equations describing the interaction between a coloured particle and gauge fields , ( [ quadriforce ] ) and ( [ isospin ] ) , to the case in which the fields are those from the julia - zee dyon solution , that are given by @xcite : @xmath77 in which the new functions are : @xmath78 and @xmath79 is a constant .
we can see that this solution has plenty of similarities with the bps monopole .
the main difference is that the dyon vector field has a non - vanishing time component , due to the existence of the non - vanishing constant @xmath79 .
it is this component that asymptotically generates an electric field that is absent in the bps monopole solution , inviting us to interpret the dyon as an entity with magnetic plus electric charge . replacing equations ( [ dyon ] ) and ( [ dyon2 ] ) in ( [ quadriforce ] ) and ( [ isospin ] )
, we reach the following equations of motion for the system : @xmath80 + \nonumber \\ & & + \frac{rh ' - h(1+k)}{\gamma^2 r^4 } \left ( \vec{i } \cdot \vec{r } \right ) \cdot \left [ \left ( \cosh{\lambda } + \gamma \sinh{\lambda } \right ) \vec{r } - \gamma \sinh{\lambda } \left ( \vec{r } \cdot \vec{v } \right ) \vec{v } \right ] + \nonumber \\ & & + \frac{k^2 - rk'-1}{\gamma r^4 } \left ( \vec{i } \cdot \vec{r } \right ) \left ( \vec{v } \times \vec{r } \right ) + \frac{k'}{\gamma r } \left ( \vec{v } \times \vec{i } \right ) \label{movdyon } \\
\dot{\vec{i } } & = & \frac{1-k}{r^2 } \left [ \vec{i } \times \left ( \vec{v } \times \vec{r } \right ) \right ] + \frac{hk\cosh{\lambda}}{r^2 } \left ( \vec{i } \times \vec{v } \right ) + \nonumber \\ & & + \frac{\left[rh'-h ( 1 + k - \tanh{\lambda})\right ] } { r^2 } \cdot \cosh{\lambda } \left ( \vec{r } \cdot \vec{v } \right ) \left ( \vec{i } \times \vec{r }
\right ) \label{movdyon2}\end{aligned}\ ] ] which bears some resemblance with the equations of motion when the interacting field is the bps monopole .
it is straightforward to note that there are no equilibrium points for this system either .
thus , as an analytical solution does not seem available , we study special cases for this system , as we have done for the case of the bps monopole .
first , one can note that the particle can develop radial motion if its initial isospin and its initial velocity point both to the radial direction ; the equations of motion resultant from these conditions are : @xmath81 where @xmath21 is still the function ( [ potencial ] ) .
we can also see that , in the newtonian limit , the equations of motion are the same as the ones referring to the bps monopole , as long as we replace the constant @xmath34 by an `` effective constant '' @xmath82 that takes into account the electric charge carried by the dyon . hence , the qualitative description of the motion is the same : no bounded orbits are allowed and the particle escapes to infinity . in the relativistic case , we can study the expressions in ( [ radial_dyon ] ) by searching for constants of motion . as in the bps monopole case , a constant that can be interpreted as the total relativistic energy of the particle
can be found and is given by : @xmath83 where @xmath84 , again . when we analyze the conditions in which return points occur , it is possible to conclude that , as in the bps monopole case , there is at most one return point .
this means that no bounded orbits are allowed even in the relativistic case .
therefore , the global behaviour of the radial orbits of systems ( [ radial ] ) and ( [ radial_dyon ] ) are very similar .
another special case of equations ( [ movdyon ] ) and ( [ movdyon2 ] ) is the asymptotic limit . making @xmath32 ,
we are led to the following expressions : @xmath85 + + \frac{\alpha}{\gamma r^3 } \left ( \vec{r } \times \vec{v } \right ) \nonumber \\ \alpha ( t ) & = & \alpha ( 0 ) \nonumber \\
\beta(t ) & = & \beta(0 ) \cos{\left ( \delta r(t ) \mbox{e}^{-\lambda } \right ) } + \rho(0 ) \sin{\left ( \delta r(t ) \mbox{e}^{-\lambda } \right ) } \nonumber \\
\rho(t ) & = & \rho(0 ) \cos{\left ( \delta r(t ) \mbox{e}^{-\lambda } \right ) } - \beta(0 ) \sin{\left ( \delta r(t ) \mbox{e}^{-\lambda } \right)}\end{aligned}\ ] ] where @xmath35 once more , as in the case of the bps monopole , the motion in the external space is not coupled to the motion in the internal space , where the isospin precessionates .
the only difference is in the frequency of the oscillations of the isospin vector , changed by the factor @xmath86 due to the dyon s electric charge .
now , we analyze the equation of motion in the external space in the newtonian limit , in which they are reduced to : @xmath87 this expression has the same similarity to the lorentz equation in eletromagnetism as described for equation ( [ assintnewton ] ) .
the difference now is the presence of the factor @xmath88 in the term referring to the electric field , due to the contribution of the dyon s electric charge .
this means that the particle is now interacting with an electric charge @xmath88 , and not with an unit one , as it was the case for the bps monopole .
three constants of motion can be found again , and are given by : @xmath89 implying that the trajectories of the particle are confined to the same conical surface of the case of the bps monopole .
the expressions for the coordinates @xmath60 and @xmath65 are the same as expressions ( [ sol : edonewton ] ) , ( [ sol : edonewton2 ] ) and ( [ phi ] ) as long as one replaces the constant @xmath34 by the `` effective constant '' @xmath90 .
the conclusions are , once again , the same as in section [ sec : asympt ] and bounded orbits occur only if @xmath91 and @xmath92 .
now , we study the first equation of ( [ assintdyon ] ) in the ultra - relativistic limit .
the expression obtained is : @xmath93 + \frac{\alpha}{\gamma r^3 } \left ( \vec{r } \times \vec{v } \right)\ ] ] one constant of motion that can be found is the total energy of the particle : @xmath94 however , as we are considering the asymptotic ( @xmath32 ) and ultra - relativistic ( @xmath95 ) limits , the first term of the previous expression is very large compared to the second , and so we can take it as constant . therefore , in such case , the velocity of the particle is a constant of the motion , as in the situation of the bps monopole .
it is straightforward to see that the total relativistic angular momentum and the module of the orbital relativistic angular momentum are also constants of motion , implying that the equations for @xmath60 and @xmath96 are the same as ( [ r_rel ] ) and ( [ phi_rel ] ) .
hence , the conclusions are the same as the ones in section [ sec : asympt ] : the orbits are unbounded and restricted to the conical surface whose axis is @xmath43 and whose half - opening angle is given by ( [ conerel ] ) .
it is possible to see that there is plenty of similarities between the two systems considered , i.e. , a particle with isospin interacting with a bps monopole and with a julia - zee dyon .
both systems present no global equilibrium points and allow the existence of radial motion consisting of unidimensional unbounded orbits . in this kind of motion , although the isospin does not play a relevant role , remaining a constant of motion not coupled to the movement in the minkowskian spacetime , its sign is important to determine the shape of the potential energy and the location of the possible return point , that indicates the direction in which the particle escapes . moreover , in the asymptotic region , considering the limits of low and high velocities , the two systems present orbits confined to a conical surface whose axis lies in the direction of the total angular momentum .
although for an intermediary velocity we can show that the orbits are not confined to a conical surface , we infer that they are confined to the region given by the two distinct cones : the one referring to the newtonian limit and the one referring to the ultra - relativistic limit . for the same initial position of the particle , it is clear that the first cone has an axis closer to the radial direction and an opening angle smaller than the cone related to the ultra - relativistic limit .
in addition , it is possible for both systems to allow stable closed orbits in the newtonian limit , if the energy of the particle is negative . in all of these asymptotic motions ,
the isospin is still not coupled to the movement in the external space , but is no longer a constant of motion , as it precessionates in the internal space .
the radial component indeed seems to play the role of an `` electric charge '' in the minkowskian spacetime and its sign is fundamental in the shape of the effective potential energy , which determines the possibility of closed orbits .
what is interesting to notice is the analogy between this asymptotic motion and the motion of a single electric charge in the presence of a magnetic monopole . as the fields of both systems , asymptotically ,
are interpreted as magnetic monopoles , one could expect to recover at least some results from classical electromagnetism if the quadriforce and isospin s evolution equations proposed were consistent .
and we indeed obtained some of these results : in the limits of high and low velocities , the particle is constrained to move over a conical surface , just as in the case of an electric charge in the presence of a monopole ( @xcite , @xcite ) .
the radial and azimuthal motions are not exactly the same in both cases , but this is not unexpected since we are dealing with two very different systems in essence ( e.g. , the symmetry group of the systems involved in this work is @xmath0 , while for electromagnetism , it is @xmath97 ) .
therefore , this simple analogy can give important clues about our choice for the equation of the quadriforce and that of the isospin s precession .
finally , we remark that , although the systems of a particle interacting with a bps monopole and a particle interacting with a julia - zee dyon seem very similar , they have important differences . in the special cases considered previously ( radial and asymptotic motion ) , these differences are revealed by the appearance of an extra factor multiplying the radial component of the isospin .
this is not unexpected , since the radial component is similar to an electric charge and the new factor carries the constant of the dyon @xmath79 that is responsible for the electric charge of the entity . | from a consistent expression for the quadriforce describing the interaction between a coloured particle and gauge fields , we investigate the relativistic motion of a particle with isospin interacting with a bps monopole and with a julia - zee dyon .
the analysis of such systems reveals the existence of unidimensional unbounded motion and asymptotic trajectories restricted to conical surfaces , which resembles the equivalent case of electromagnetism . |
black hole thermodynamics has a variety of applications , from the insights of quantum gravity gedankenexperiments to practical calculations in heavy ion collisions and condensed matter physics . and there are just as many formalisms for describing the thermodynamic behavior : the classical four laws of black hole mechanics , quantum fields on curved backgrounds , microstate counting by virtue of the cardy formula , and the euclidean path integral formulation , to name a few .
each approach has its benefits , but the path integral approach has proven especially useful for practical calculations at 0- and 1-loop level , particularly in the context of gauge / gravity dualities .
this is due , in part , to its direct connection with the classical formulation of the gravitational theory . given an action @xmath0 for the theory , every relevant field configuration contributes with weight @xmath1 .
of course , the identification and enumeration of the ` relevant field configurations ' may present a challenge .
already in the early days of path integrals it was discovered that the most relevant class of paths behave like the ` weierstrass monster ' : they are continuous but not differentiable at any point . in a simple system , like a point particle in certain potentials ,
the contributions from these paths can be accounted for .
but the prospect of summing contributions from non - differentiable field configurations seems overwhelming for a theory as complex as gravity .
instead , calculations that employ the gravitational path integral will often focus exclusively on the contributions from geometries that satisfy some smoothness conditions .
this assumption may be justified by physically reasonable results , but the fact remains that smooth metrics account for only a fraction of the support of the path integral measure .
the alternative is to require only continuity of the metric when performing path integral calculations , while relaxing the requirement of differentiability . in this paper
we make some modest progress in this direction by including metrics with conical singularities in the euclidean path integral for a number of gravitational theories .
specifically , we consider the euclidean partition function for a canonical ensemble defined inside a finite cavity .
field configurations in the ensemble satisfy certain boundary conditions at the wall of the cavity , where the system is coupled to an external thermal reservoir . in some cases ( theories with asymptotically anti - de sitter boundary conditions , for instance )
the walls of the box can be removed to an asymptotic region without compromising the existence of the ensemble , but this is not always possible . usually one computes the partition function for this ensemble by summing contributions from metrics that are regular everywhere .
we will relax this condition and include certain configurations with a conical singularity .
this ensemble could be referred to as the _ conical ensemble _ , to distinguish it from the usual canonical ensemble .
we then pose the following questions : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ how do configurations with conical singularities contribute to the partition function ?
can the ground state of the ensemble have a conical singularity ? _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ these questions can be answered quite generally for two - dimensional dilaton gravity .
this class of models includes the spherically symmetric reduction of higher - dimensional theories that admit schwarzschild , schwarzschild - ads , reissner - nordstrm , and btz black holes as solutions , as well as target space actions associated with certain string theory black holes .
how are conical singularities incorporated into the ensemble ? in the semiclassical approximation the partition function is dominated by solutions of the classical equations of motion , with sub - leading contributions coming from smooth fluctuations of the fields around these configurations .
we now wish to take into account geometries that are regular everywhere except for a single point .
the dominant contributions would seem to come from configurations that `` almost '' extremize the action : they satisfy the equations of motion everywhere except for the singular point , similar to the ` off - shell black holes ' considered by carlip and teitelboim @xcite .
the assumptions of our framework then require this point to sit either at the center of the cavity , or the horizon of a black hole .
from the point of view of a higher dimensional model , a conical singularity at any other point other than the center of the cavity would not be consistent with the spherically symmetric reduction of the action . in the context of the two - dimensional dilaton gravity ,
the fact that the configurations satisfy the equations of motion everywhere except a single point implies the existence of a certain killing vector that forces the singularity to sit at the center of the cavity .
thus , the dominant contributions to the partition function from configurations with a conical singularity correspond to solutions of the equations of motion that break down at the horizon , where there is a delta - function singularity in the curvature . in the canonical ensemble
, one finds that the stable ground state of the system is typically either ` hot empty space ' , a regular black hole , or some superposition of the two , depending on the boundary conditions . even when it is not the ground state a black hole may exist for certain boundary conditions as a local minimum of the free energy , stable against small thermodynamic fluctuations .
we find that the inclusion of conical singularities in the ensemble does not change either of these statements .
when the boundary conditions allow a thermodynamically stable black hole in the canonical ensemble , the addition of a small conical defect a conical singularity with deficit or surplus angle @xmath2 may result in a lower internal energy or a higher density of states . in other words , small conical defects may be energetically or entropically favored over their smooth counterparts in the conical ensemble .
@xmath3 these results are somewhat surprising because they seem to contradict our intuition that smooth black holes should be the most stable configurations . however , and are not possible simultaneously , and in fact the presence of a small conical defect always increases the free energy : @xmath4 we conclude that black holes that are stable against gaussian fluctuations in the canonical ensemble are also stable against the nucleation of a small conical defect .
this perturbative stability generalizes to a non - perturbative statement about the ground state of the system . with a few caveats
, it appears that the ground state of the conical ensemble is always a smooth field configuration .
in fact , the relationship between the ground state and the boundary conditions is the same as in the canonical ensemble .
we provide general arguments as well as several examples demonstrating these features . in the semiclassical approximation
the partition function is dominated by the ground state of the system , which is a smooth geometry .
corrections to this leading behavior from small , smooth fluctuations around the ground state are well - understood and can be evaluated using a variety of techniques . since configurations with a conical singularity
do not dominate the partition function , it is natural to ask how their contributions compare to the corrections from smooth fluctuations .
the action is generally a complicated function of the fields , so contributions from conical singularities can not be evaluated analytically except in special cases . however , it is possible to evaluate the contributions numerically , and in the semiclassical limit the results are approximated to high precision by relatively simple functions of the boundary conditions that define the ensemble .
when the ground state is a black hole , we find that the contributions to the partition function from configurations with a conical singularity are comparable to the contributions from gaussian fluctuations .
this suggests that , even in the semiclassical approximation , non - smooth field configurations make important sub - leading contributions to thermodynamic quantities .
it is important to point out that we view these contributions as logically distinct from the `` mass fluctuations '' that are already studied in the literature , see e.g. @xcite , or from 1-loop quantum corrections derived by taking into account conical defects @xcite .
indeed , the reader may wonder whether the configurations we consider are not already included in those calculations . in the canonical ensemble
the proper temperature is held fixed at the wall of the cavity , and hence black holes with a regular horizon are only present in the ensemble for isolated values of the mass ( if they are present at all ) .
shifting the black hole mass away from these values necessarily introduces a conical singularity at the horizon , and this must be addressed before the role of these configurations in the ensemble can be understood .
thus , we assume that `` mass fluctuations '' in previous calculations like @xcite refer only to small , smooth fluctuations of the fields around a black hole background with a fixed horizon , as opposed to an actual variation in the mass of the black hole .
on the other hand , 1-loop corrections depend explicitly on the precise matter content coupled to the gravitational theory , so they are not exclusively an intrinsic property of the black hole .
see for instance @xcite , eqs .
( 1.1 ) , ( 1.2 ) , and refs . therein .
thus , our approach of considering non - smooth geometric configurations that are on - shell everywhere except at the horizon is logically distinct form these two approaches . nevertheless , as we will show
, the leading corrections to quantities like the entropy take essentially the same form as they do in other approaches .
this paper is organized as follows . in section [ sec:1 ]
we recall some basic results of two - dimensional dilaton gravity and black hole thermodynamics in the canonical ensemble . in section [ sec:2 ]
we study black holes with a conical defect and evaluate the conical ensemble partition function in the semiclassical approximation . in section [ sec:2a ]
we address general features of thermodynamical observables , discuss stability issues , and give approximate expressions for the contributions to the partition function . in section [ sec:3 ]
we provide several explicit examples , and in section [ sec:4 ] we point to applications and open problems . before proceeding we point out a few important conventions , which are the same as @xcite in most cases .
euclidean signature is employed throughout .
nevertheless , terms appropriate to lorentzian signature such as ` minkowski ' and ` horizon ' are used when the meaning is clear .
we use natural units with @xmath5 , and the dimensionless newton s constant in two dimensions is set to @xmath6 .
these constants are restored in certain expressions , when necessary .
in this section we recapitulate some of the main results of @xcite . a reader familiar with these results and
the notation may skip this section .
in this paper we study black hole ( bh ) thermodynamics in two - dimensional models of dilaton gravity .
dilaton gravity in two dimensions is conventionally described by the euclidean action @xmath7 = & \,\ , - \frac{1}{2 } \int_{{\mathcal{m } } } { \negthickspace}{\negthickspace}{\mathrm{d}}^{\,2}x \,\sqrt{g}\ , \left ( \,x\,r - u(x)\,\left(\nabla x\right)^2 - 2 \ , v(x ) \ , \right ) \\ \nonumber & \,\ , - \int_{{\partial { \mathcal{m } } } } { { \negthickspace}{\negthickspace}{\negthickspace}}{\mathrm{d}}x \,\sqrt{\gamma } \ , x\,k + \int_{{\partial { \mathcal{m } } } } { { \negthickspace}{\negthickspace}{\negthickspace}}{\mathrm{d}}x \,\sqrt{\gamma } \ , \sqrt{w(x)e^{-q(x ) } } ~.\end{aligned}\ ] ] the dilaton @xmath8 is defined in terms of its coupling to the ricci scalar , which takes the form @xmath9 .
different models are distinguished by the kinetic and potential functions @xmath10 and @xmath11 , cf .
e.g. @xcite for reviews .
the bulk terms in the first line of are supplemented by boundary terms in the second line .
the first boundary term is the analog of the gibbons - hawking - york surface integral @xcite , where @xmath12 is the induced metric on the boundary and @xmath13 is the trace of the extrinsic curvature . including this term in the action ensures a dirichlet boundary problem .
the second boundary term is the holographic counterterm derived in @xcite .
it ensures a well - defined variational principle , so that the first variation of the action vanishes on solutions of the equations of motion for all variations that preserve the boundary conditions and @xmath14 appearing in the counterterm is the supergravity pre - potential @xcite . ] .
the functions @xmath15 and @xmath14 , which depend on @xmath10 and @xmath11 , are defined below .
the equations of motion are obtained by extremizing the action with respect to small variations of the fields that preserve the boundary conditions .
this yields @xmath16 solutions of these equations always possess at least one killing vector @xmath17 with orbits that are curves of constant @xmath8 @xcite .
fixing the gauge so that the metric is diagonal , the solutions take the form @xmath18 with @xmath19 the solutions depend on an integration constant @xmath20 , as well as two model - dependent functions @xmath14 and @xmath15 that are given by integrals of @xmath10 and @xmath11 @xmath21 the integrals are evaluated at @xmath8 , with constants of integration @xmath22 and @xmath23 .
notice that @xmath23 and @xmath20 contribute to @xmath24 in the same manner ; together they represent a single parameter that has been partially incorporated into the definition of @xmath15 . by definition
they transform as @xmath25 and @xmath26 under the shift @xmath27 .
this ensures that the functions and transform homogeneously , allowing @xmath22 to be absorbed into a rescaling of the coordinates but not on @xmath22 . ] .
therefore , the solution depends on a single constant @xmath28 . with an appropriate choice of @xmath23 we can restrict @xmath20 to take values in the range @xmath29 for physical solutions . as evident from the norm of the killing vector @xmath17 is @xmath30 .
if it vanishes we encounter a killing horizon .
solutions with @xmath31 , which exhibit horizons , will be referred to as bhs .
if the function @xmath11 happens to have a zero at @xmath32 then there is a second , inequivalent family of solutions that also have the form .
the dilaton and metric for these solutions are given by @xmath33 where @xmath34 and @xmath35 are arbitrary constants .
in most applications these solutions , which are characterized by a constant dilaton and ricci scalar , are not relevant .
we will generally ignore them , so references to `` generic solutions '' or `` all solutions '' should be understood to mean the solutions and , parametrized by the mass @xmath20 . in the models we consider , solutions have a non - negative metric function @xmath24 over a semi - infinite interval @xmath36 with the lower end of this interval corresponding to either the origin or a horizon , and the upper end corresponding to the asymptotic region of the space - time .
at the upper end of the interval the function @xmath15 generally diverges @xmath37 so the asymptotic behavior of the metric is characterized by the solution with @xmath38 .
if the metric function is strictly positive then the lower end of the interval is just the value of the dilaton at the origin .
but if @xmath24 vanishes for some value of the dilaton , @xmath39 , then the lower bound is a killing horizon . assuming that the function @xmath40 is non - zero for finite values of @xmath8 , the location of the horizon is related to the parameter @xmath20 by @xmath41 if this condition admits multiple solutions then @xmath42 is always taken to be the outermost horizon , so that @xmath43 for @xmath44 . for a field configuration to extremize the action it should satisfy the equations of motion at all points , and this imposes certain differentiability conditions on solutions .
in particular , for solutions with @xmath45 the horizon must be a regular point of the geometry .
this fixes the periodicity @xmath46 of the euclidean time , which is given by @xcite @xmath47 the inverse periodicity is related to the surface gravity of the bh by @xmath48 . in an asymptotically flat space - time @xmath49 is also the temperature measured by an observer at infinity , so we denote this quantity by @xmath50 @xmath51 this slight abuse of notation should not be confused with the proper local temperature @xmath52 , which is related to @xmath49 by a dilaton - dependent ` tolman factor'@xcite @xmath53 the proper temperature at infinity coincides with only if @xmath54 as @xmath55 . a solution with @xmath38 has no horizon , and is therefore regular everywhere without having to place any conditions on the period of the euclidean time .
however , as we will see below , the boundary conditions of the canonical ensemble determine a unique , non - zero value for the otherwise arbitrary period .
we will therefore refer to this solution , which does not contain a black hole but has a non - zero temperature , as ` hot empty space ' ( he s ) . to describe a consistent bh thermodynamics
we must specify an ensemble and construct the appropriate partition function .
motivated by york s path integral formulation of the canonical ensemble @xcite , we introduce an upper bound @xmath56 on the interval .
this constrains the dilaton to a ` cavity ' @xmath57 whose wall is the dilaton isosurface @xmath58 .
quantities evaluated at the wall or with an explicit dependence on @xmath56 will carry a subscript ` @xmath59 ' .
boundary conditions for the canonical ensemble are imposed by coupling the system to a thermal reservoir , which fixes a dilaton charge @xmath60 and the proper local temperature @xmath61 at the wall of the cavity , one can construct a current that yields that function as its conserved charge . for simplicity ,
we take the dilaton charge at the wall to be @xmath62 , and refer to this boundary condition henceforth as fixing @xmath56 .
a detailed discussion can be found in @xcite , or in section 3.1 of @xcite . ] .
it is convenient to think of the boundary condition on the temperature as fixing the proper local period of the euclidean time , which is just the inverse @xmath63 of the proper local temperature .
the proper local period is related to the period @xmath46 by @xmath64 when combined with the smoothness condition this becomes @xmath65 this model - dependent ( and often complicated ) equation , which may or may not have solutions @xmath66 , determines whether there are smooth black holes in the ensemble for given boundary conditions @xmath67 and @xmath56 .
not all solutions of this equation are relevant : the upper bound on the dilaton implies that only solutions with @xmath68 ` fit ' inside the cavity .
thus , any solutions @xmath20 of that lie in the range @xmath69 are elements of the canonical ensemble , where @xmath70 corresponds to a black hole with horizon located at the wall of the cavity .
one solution that almost always appears in the canonical ensemble is he s ( @xmath38 ) with period fixed by the boundary condition @xmath71 in most models the he s solution dominates the ensemble for at least some range of boundary conditions @xcite . with these boundary conditions
the partition function of the canonical ensemble is given by the euclidean path integral @xmath72 \right ) ~.\ ] ] for now we will take the path integral to include all smooth spaces @xmath73 and dilaton configurations @xmath8 that satisfy the boundary conditions , but we will relax the smoothness requirement in the next section . in the semi - classical limit the dominant contribution to
the euclidean path integral comes from the minimum of the action .
the minimum is of course a stationary point of the action either a black hole with @xmath31 satisfying ( if such a solution exists ) , or he s with period .
so the action for smooth field configurations near the minimum can be written as @xmath74 = & \,\ , { \gamma}[g_{\textrm{\tiny min}},x_{\textrm{\tiny min } } ] + { \delta}{\gamma}[g_{\textrm{\tiny min } } , x_{\textrm{\tiny min } } ; \delta g , \delta x ] \\
\nonumber & \,\ , + \frac12 \,{\delta}^2{\gamma}[g_{\textrm{\tiny min } } , x_{\textrm{\tiny min } } ; \delta g , \delta x ] + \ldots \end{aligned}\ ] ] where @xmath75 and @xmath76 are the linear and quadratic terms in the taylor expansion .
in @xcite it was shown that the leading term is finite and the linear term vanishes for all field variations @xmath77 , @xmath78 consistent with the boundary conditions .
since the solution is assumed to be a minimum ( as opposed to a saddle point ) the quadratic term is positive definite and the semi - classical approximation of the path integral is given by @xmath79\right ) \times ( \text{quadratic corrections})~,\ ] ] where the second factor comes from performing the ( gaussian ) integral over the quadratic terms in .
the partition function exhibits qualitatively different behavior depending on whether the minimum of the action is he s or a black hole . for a particular model
the ground state can be readily determined from the values of the boundary conditions : one simply identifies the solutions of the equations of motion that belong to the ensemble , and then determines which solution has the smallest action for the given values @xmath67 and @xmath56 .
evaluating the action for the solutions and gives @xmath80 which is bounded below for finite @xmath56 .
for he s with @xmath81 this becomes @xmath82 where @xmath83 is the value of the dilaton at the origin ( in most of the models we consider , @xmath84 ) . if is less than for all relevant solutions of , then he s is the ground state of the ensemble . on the other hand , if there is a solution of such that is less than , then the ground state of the ensemble is a black hole .
if the values of @xmath67 and @xmath56 are changed then the ground state of the ensemble may change as well , in which case the system will undergo a phase transition involving the nucleation of the new ground state either a stable black hole or he s . assuming that there is a single minimum of the action that minimize the action .
for instance , it may be possible to tune @xmath67 and @xmath56 so that he s and a black hole both have the same action .
] , the dominant semiclassical contribution to the free energy @xmath85 is given by @xmath86 from this result it is possible to derive all thermodynamical properties of interest , like the entropy and internal energy @xmath87 a comprehensive discussion of thermodynamical properties is provided in @xcite .
a key assumption in the derivation of is that the quadratic term in must be positive definite .
this is just the thermodynamic stability condition that the ground state have positive specific heat at constant dilaton charge .
the specific heat at constant @xmath56 is @xmath88 which yields @xmath89 thus , given boundary conditions @xmath56 and @xmath61 , a canonical ensemble dominated by a black hole exists if the minimum of the action is a solution @xmath90 of , and @xmath91 so that the specific heat is positive .
an important point is that for some theories this inequality can only be realized for finite @xmath56 .
a theory whose boundary conditions and solutions have @xmath92 will not have positive specific heat as @xmath93 , since @xmath94 in this limit . in that case a finite cavity is required for the existence of the canonical ensemble .
the classic example of this phenomenon is the schwarzschild black hole in a theory with asymptotically flat boundary conditions @xcite . on the other hand ,
the right - hand side of is strictly negative , so a finite cavity is not required for a theory with boundary conditions and solutions such that @xmath95 . in that case
the specific heat remains positive as @xmath93 .
if the action of the black hole is less than the action for he s in this limit , then there is a canonical ensemble with black hole ground state that does not require an external thermal reservoir .
the expression for the specific heat will play an important role in later sections .
this concludes our review of thermodynamics in the euclidean path integral formalism for two - dimensional dilaton gravity . in the rest of the paper
we weaken the assumption of differentiability by considering continuous metrics that no longer satisfy the condition .
let us now reconsider the euclidean partition function . in the previous section we included contributions from smooth field configurations that satisfy the boundary conditions .
then the leading term in the semi - classical approximation of the canonical partition function is @xmath96 with @xmath20 being the mass of the smooth solution that minimizes the euclidean action and @xmath97 the on - shell action .
we assume as before that the absolute minimum of the action occurs for a single value of @xmath20 ; otherwise contains a sum over all values of @xmath20 that minimize the action .
we may rewrite as @xmath98 where @xmath99 runs over the physically allowed values of the mass subject to the condition @xmath100 and the delta - function picks out the value @xmath101 of the smooth solution that minimizes the action .
now suppose that we enlarge the class of field configurations that contribute to @xmath102 by relaxing the assumption of smoothness . instead of imposing the condition for the period ,
we allow for metrics that are continuous but not differentiable at the center of the cavity . in the semiclassical limit
, the largest contributions to the partition function from this sector are expected to come from configurations that ` almost ' meet the conditions for stationary points of the action : they satisfy the equations of motion at all points except for the horizon , where there is a conical singularity . assuming the action is well - defined for these backgrounds , the contributions to @xmath102 take the form without the delta - function as an integral over internal energy @xmath103 with the weight given by @xmath104 . ]
@xmath105 in the rest of this section we will consider the properties of the ` conical defect ' black holes that contribute to this integral , and evaluate the action @xmath106 that appears in the exponent .
black hole field configurations that satisfy the boundary conditions but _ not _ the smoothness condition have the same general form as before @xmath107 with period @xmath108 , and the functions @xmath8 and @xmath109 given by @xmath110 the location of the horizon , @xmath100 , is determined as before : @xmath111 we will sometimes denote a function s dependence on @xmath112 ( as opposed to the values @xmath20 that satisfy , which typically comprise a discrete set ) with a ` hat ' , abbreviating @xmath113 as @xmath114 and @xmath115 as @xmath116 . by assumption , most values of the parameter @xmath112 do not correspond to smooth black holes the condition is not satisfied .
this means that the periodicity of the euclidean time is not equal to the period required for regularity at the horizon . instead , the period @xmath117 is determined by the boundary condition and the parameter @xmath118 , @xmath119 in other words , the period @xmath120 does not agree with @xmath121 , defined as @xmath122 as a result , these spaces exhibit a a conical singularity .
the deficit ( or surplus ) angle @xmath123 associated with the defect is @xmath124 where @xmath125 . if @xmath126 then @xmath127 is positive , and it represents a true deficit in the period of euclidean time
otherwise , if @xmath127 is negative there is a surplus in the period . for convenience we will always refer to @xmath123 as the ` deficit angle ' , though it may be positive or negative .
an important distinction between spaces with a conical defect and smooth solutions of the equations of motion is that @xmath112 is a continuous parameter that is independent of the boundary conditions @xmath67 and @xmath56 .
the only conditions on @xmath112 are that it should lie in the range associated with physical solutions , @xmath128 , and the horizon should fit inside the cavity , @xmath100 .
the discrete set of masses @xmath20 that correspond to smooth solutions are determined by the condition , which implies a ( potentially complicated ) dependence on @xmath67 and @xmath56 . in the presence of a conical singularity
the action must be evaluated carefully , taking into account the behavior of the metric at the horizon .
this is accomplished using a smoothing procedure that regulates the defect @xcite .
when the action is evaluated for a black hole with conical singularity the result is independent of the details of the smoothing .
this suggests that the action , without any modifications , is appropriate for weighting contributions to the partition function from these spaces , as in . for our purposes , a conical singularity at a point @xmath129 on the interior of @xmath130
can be thought of as introducing a delta - function in the curvature , in the sense that the integral of the ricci scalar over @xmath130 is @xmath131 where @xmath127 is the deficit angle and @xmath132 is the manifold @xmath130 with the point @xmath129 removed .
the spaces described in the previous section have a conical defect at the horizon @xmath133 , so we can write the action as the usual functional on @xmath134 the manifold @xmath130 with the singular point removed plus the contribution from the defect @xmath135 = & \,\ , - \frac{1}{2}\,\int_{\hat{\mathcal{m } } } { \negthickspace}{\mathrm{d}}^{\,2}x \,\sqrt{g}\ , \left [ x\,r - u(x)\,\left(\nabla x\right)^2 - 2 \ , v(x ) \right ] \\
\nonumber & \,\ , - \int_{{\partial { \mathcal{m } } } } { { \negthickspace}{\negthickspace}{\negthickspace}}{\mathrm{d}}x \,\sqrt{\gamma } \ , x\,k + \int_{{\partial { \mathcal{m } } } } { { \negthickspace}{\negthickspace}{\negthickspace}}{\mathrm{d}}x \,\sqrt{\gamma } \ , \sqrt{w(x)e^{-q(x ) } } - \hat{x}_h\,{\alpha}~.\end{aligned}\ ] ] for @xmath136 and constant dilaton the resulting functional of the metric is , up to an overall factor of @xmath137 , the gauss - bonnet formula for a compact euclidean manifold @xmath130 with boundary @xmath138 and a deficit angle @xmath123 . with the results above the on - shell action for a black hole with a conical singularity at the horizon
is @xmath139 the free energy , which encodes all thermodynamical properties of interest , is then given by @xmath140 when considering the role of black holes with a conical singularity in the ensemble it is useful to compare to the action for a smooth solution of the equations of motion .
the difference between the actions is given by @xmath141 this result will be useful in the next section .
we investigate now the thermodynamical properties of field configurations with a conical defect , and compare their role in the conical ensemble to that of smooth solutions . for
now we assume that the ensemble contains among the smooth field configurations a black hole with mass @xmath20 , horizon @xmath142 , and positive specific heat .
thus , the black hole is at least a local minimum of the action among smooth spaces , though it need not be the absolute minimum of the action .
it is useful to define a quantity @xmath143 that relates the dilaton at the conical defect to the dilaton at the horizon of the smooth solution @xmath144 after deriving expressions for the entropy and internal energy , we consider the case @xmath145 where the thermodynamic properties of conical defects can be analyzed perturbatively . later in this section
we derive results that are valid non - perturbatively , in particular concerning the stability of smooth configurations and the ground state of the ensemble . in two - dimensional dilaton gravity
the entropy of smooth black holes takes the universal form .
this result , which is independent of both the details of the theory and the size of the cavity , generalizes to black holes with a conical singularity @xmath146 in terms of the parameter @xmath147 this becomes @xmath148 and we see that the entropy may either be greater than or less than the entropy of the smooth black hole , depending on the sign of @xmath147 . the internal energy is related to the entropy , temperature , and free energy by @xmath149 .
the subscript ` c ' is retained here to emphasize that , unlike the entropy , the internal energy depends explicitly on the size of the cavity . applying the results for the entropy and the free energy gives @xmath150 like the entropy , the internal energy of a black hole with a conical singularity may be either greater than or less than that of the smooth black hole . based on the results above ,
a black hole with a conical singularity can have higher entropy or lower internal energy than a smooth black hole .
however , to determine whether these black holes are favored in the conical ensemble one must consider the free energy . we will first address this for small @xmath143 ; i.e. , @xmath151 . in this
limit the deficit angle is @xmath152 so positive @xmath143 corresponds to a surplus , and negative @xmath143 represents a deficit .
the expression for the entropy is linear in @xmath113 , so no expansions are needed when @xmath143 is small . on the other hand ,
the internal energy is a non - linear and potentially complicated function of @xmath113 .
expanding for small @xmath143 gives @xmath153 so the internal energy of the conical defect may be greater or less than the internal energy of the smooth black hole , and at leading order this is controlled by the sign of @xmath143 .
notice , however , that the term at order @xmath154 is strictly positive .
this is crucially important when we consider the free energy @xmath155 . using and the expansion we obtain @xmath156 expressed in terms of the deficit angle this is @xmath157 thus , the free energy of a smooth black hole with @xmath158 is always smaller than the free energy of a nearby ( @xmath159 ) black hole with conical singularity . in terms of the internal energy and entropy
this implies @xmath160 in other words , the presence of a small conical defect can lower the internal energy compared to a smooth black hole , but the corresponding decrease in the density of states prevents the ensemble from favoring such configurations .
likewise , a conical defect can have a larger entropy than a smooth black hole , but the cost in internal energy is too high for these configurations to be favored by the ensemble .
the previous section considered black holes with a small conical defect . in this section
the assumption @xmath161 is dropped , which means that quantities like @xmath162 can not be evaluated perturbatively . nevertheless , it is still possible to show that the minimum of the action does not have a conical defect . if the absolute minimum of the action among smooth spaces is a solution with mass @xmath20 , then @xmath163 for any @xmath112 with a conical singularity .
the ensemble always has a smooth ground state that is stable against decay into a space with conical defect .
first let us illustrate our reasoning with a simple class of examples : theories that allow the @xmath93 limit .
the existence of the ensemble in this limit is addressed in the next section ; for now let us assume that we are working with a model where taking @xmath93 is allowed . then as the cavity wall is removed becomes @xmath164 the condition @xmath165 becomes a convexity condition on the function @xmath166 .
an even stronger statement is obtained by considering extrema of @xmath167 .
for very large @xmath56 the condition @xmath168 simplifies to @xmath169 but this implies that any extremum of @xmath167 has to have the same periodicity as the configuration without conical defect . while more than one such extremum may exist ( cf .
the discussion about how to extract the mass @xmath20 from @xmath61 and @xmath56 in @xcite ) , none of them exhibits a conical singularity . for finite values of @xmath56
it is easier to work directly with the action for a space with conical singularity . extremizing the action with respect
to @xmath112 gives @xmath170 which means that extrema occur at precisely those values of @xmath112 that correspond to smooth solutions of the equations of motion .
of course , it is possible that these extrema are local , and the absolute minimum of the action occurs at one of the endpoints of the interval @xmath171 .
indeed , in most models there is a range of boundary conditions where the minimum of the action occurs at the lower limit .
but @xmath172 is just he s , which is a smooth solution of the equations of motion .
the other possibility that the absolute minimum of the action occurs at @xmath173 can be ruled out quite generally .
consider the derivative @xmath174 as @xmath175 from below @xmath176 for non - zero @xmath67 the first term is positive and diverges as @xmath177 . unless @xmath178 happens to have a zero at @xmath56 , the first term dominates and the action is _ increasing _ as @xmath112 approaches @xmath179 .
we conclude that the action is always minimized by a smooth solution of the equations of motion : either he s or a smooth black hole .
note that it is possible for a conical singularity to have a smaller action than a smooth black hole , as long as that black hole is not the ground state of the ensemble .
this includes black holes that are thermodynamically stable ( @xmath180 ) , but only a local minimum of the action . in that case
there will necessarily be conical singularities close to the ground state that have a smaller action than any local minimum .
but a smooth black hole ( or he s , for that matter ) will never tunnel quantum mechanically to a final configuration with a defect , because the ground state of the ensemble is necessarily smooth .
the discussion up to this point has involved generic solutions of the equations of motion , but neglected the constant dilaton vacua ( cdv ) - that may exist for some dilaton gravity models .
these isolated solutions occupy a different superselection sector of the theory , so there is no perturbative channel for a bh with or without a deficit angle to decay into a cdv . however , in cases where the boundary conditions happen to coincide with a zero of the dilaton potential
, @xmath181 , tunneling between the two types of solutions is possible .
a detailed discussion can be found in @xcite . since we have extended the class of bh solutions to configurations with a deficit angle , it is appropriate to do the same for cdvs .
here we discuss these solutions and evaluate their free energy .
the on - shell action can be calculated using , which gives @xmath182 ( if we drop the assumption that spacetime is topologically a disk the first term in is multiplied by the euler characteristic of the manifold . )
the difference between this action and the action for a smooth cdv solution is @xmath183 which always vanishes because both configurations satisfy the same boundary conditions @xmath184 therefore , all cdv solutions with given @xmath83 and @xmath185 have the same free energy .
it follows that the regular cdv solution is only marginally stable against decay into a cdv with conical defect , and vice - versa .
the dominant contributions to the semiclassical partition function from spaces with a conical singularity are given by an integral like .
but for systems with a finite cavity the measure in this integral should be treated more carefully . in the canonical ensemble
the partition function is expressed as a sum or integral over different internal energies of the system , weighted by the density of states @xmath186 and the boltzmann factor @xmath187 .
the internal energy for an ensemble with finite @xmath56 is given by , which suggests that the appropriate measure is proportional to @xmath188 rather than @xmath189 .
thus , the semiclassical approximation for the partition function , including contributions from spaces with a conical singularity , is @xmath190 the additional factor of @xmath191 in the measure proves to be relevant when computing sub - leading corrections to thermodynamical quantities like the entropy .
the functional form of the integrand in almost always prevents the direct evaluation of this integral in closed form , but standard semiclassical approximation techniques prove to be very accurate when compared with numerical results . for some theories the conical ensemble exists even as the system is decoupled from the external thermal reservoir .
taking the @xmath192 limit , or ` removing the cavity wall ' , usually implies @xmath94 , and in this case the integral simplifies . provided the limit @xmath93 commutes with the integral over @xmath112 ,
the contributions to the semiclassical partition function becomes @xmath193 assuming there are no obstructions , we can use @xmath194 to convert this into an integral over @xmath133 @xmath195 where the lower bound is set by the condition @xmath196 , and we have absorbed the factor @xmath197 into the normalization of @xmath198 in the measure in .
the point is that this factor is important for finite @xmath56 , but it becomes state - independent , and hence irrelevant , in the @xmath93 limit . ] . before going further
it is important to ask : if @xmath93 , what is being fixed in defining the ensemble ?
if @xmath199 is finite and non - zero in this limit then we may continue to express the action as a function of the same @xmath67 that is held fixed at finite @xmath56 .
but if @xmath199 diverges then we must take @xmath200 while keeping the ratio @xmath201 finite . in either case , the boundary conditions of the ensemble are specified by fixing a finite , non - zero value for the quantity @xmath202 in other words , when @xmath93 the ensemble is defined by fixing the value of the period , rather than the proper local period at the cavity wall .
the contributions to the partition function from conical singularities are then given by @xmath203 which can also be expressed in the familiar form @xmath204 of course , the ensemble only exists if this integral is defined , which requires that @xmath15 grows sufficiently fast at large values of @xmath8 , @xmath205 if this condition is satisfied then the ensemble exists . for example , the @xmath93 limit is not defined for the schwarzschild model , which has @xmath206 , but it is defined for the jackiw - teitelboim model , which has @xmath207 . if the large @xmath8 behavior of @xmath15 is linear , so that @xmath208 for large @xmath8 , then the ensemble exists only if @xmath209 , which corresponds to a hagedorn temperature @xmath210 .
this is especially relevant for the stringy black holes considered in section [ subsec : stringy ] . in most cases
the integral is easier to work with than the integral , though approximation methods or numerical techniques are usually still required .
the behavior of this integral depends on the ground state of the system , which is determined as in the finite @xmath56 case .
the condition for extremizing the action in the @xmath93 limit is @xmath211 where @xmath212 is the action in the exponent of .
as before , this is the usual smoothness condition , so the ground state of the system will either be a smooth black hole or the @xmath213 he s solution case , where the upper limit @xmath179 had to be considered when determining the ground state , the upper limit @xmath214 is ruled out by the condition . ] .
a solution @xmath215 of is a minimum if @xmath216 which is equivalent to the @xmath93 limit of the condition for positivity of the specific heat .
if a minimum exists , it must be compared to the action of he s , @xmath217 , to determine the ground state .
thus , the ground state of the ensemble is a smooth black hole if there is a solution of that satisfies and @xmath218 otherwise , the ground state is he s .
when the ground state of the ensemble is a black hole , the integral is comparable to contributions to the partition function from smooth gaussian fluctuations . to see why this is the case ,
define a new variable @xmath219 where @xmath42 is the dilaton at the horizon for the black hole ground state . in the semiclassical approximation the main contributions to come from configurations close to the ground state ,
so the integral may be approximated as @xmath220 where @xmath221 is the @xmath93 limit of the specific heat .
evaluating the integral gives @xmath222 this is comparable to the contributions from _ smooth _ gaussian fluctuations around the black hole ground state , because in both cases the coefficient of the quadratic term in the expansion of the action around the minimum is proportional to @xmath223 .
if the ground state of the ensemble is he s , then there are two possible approximations for the integral . as before , configurations close to the ground state dominate the integral in the semiclassical approximation .
but their contributions depend on the behavior of the function @xmath224 in this region .
we will assume , for convenience , that @xmath225
. then if the ground state is he s , the conditions @xmath226 and @xmath227 imply that @xmath228 near @xmath229 , with @xmath230 . for @xmath231 ,
the @xmath232 term dominates the action in this region , and the integral is well - approximated as @xmath233 on the other hand , if @xmath224 is approximately linear ( @xmath234 ) near @xmath235 then both terms in the action are relevant , and is approximated by @xmath236 in both cases the contributions to the partition function are generally much smaller than other corrections ( for instance , from radiation ) for a he s ground state . though the analysis is a bit more complicated for finite @xmath56 , the same approach yields reliable approximations for .
when the ground state of the ensemble is a black hole , expanding the action around its minimum gives @xmath237 where @xmath219 , and @xmath238 is the specific heat at finite @xmath56 .
the semiclassical limit implies that the integral is dominated by configurations near the minimum , and their contributions may be approximated as @xmath239 this gives essentially the same result as @xmath240 but expressed in terms of the relevant quantities evaluated at @xmath56 .
when the ground state is he s , the analysis is very similar to the @xmath93 case .
configurations near @xmath213 dominate the integral , and depending on the behavior of @xmath224 in this region is approximated by either or , with @xmath241 replaced by @xmath242 . for comparison with other approaches that calculate corrections to free energy and entropy it is useful to represent the results above as @xmath243
then the entropy takes the form @xmath244 , where @xmath245 is the contribution from the leading term in the free energy , and @xmath246 the second line gives the leading behavior of @xmath247 in our semiclassical calculations , which takes the same form as corrections from thermal fluctuations @xcite
. these results also apply to ensembles with @xmath93 , after making the appropriate replacements of quantities evaluated at the cavity wall .
so far the discussion has allowed for arbitrary functions @xmath10 and @xmath11 in the action . in this section
we apply our results to specific models , discussing their thermodynamic properties and determining the leading contributions to the partition function from configurations with conical singularities .
the schwarzschild models , which belong to the so - called ` @xmath248-family ' of dilaton gravities @xcite , are motivated by a spherically symmetric reduction of gravity with asymptotically flat boundary conditions from @xmath249 dimensions down to two dimensions .
the functions @xmath15 and @xmath40 for these models take the form @xmath250 the constant @xmath251 is given by @xmath252 where @xmath253 is the @xmath254-dimensional newton s constant and @xmath255 is the area of the unit sphere @xmath256 .
it will be convenient to retain factors of @xmath251 , even though they could be absorbed into a rescaling of the coordinates .
since the growth of @xmath15 is sub - linear for large @xmath8 , the existence condition implies that the @xmath93 limit is not possible for the schwarzschild model .
thus , in order to work in the canonical ensemble one must couple the system to a heat bath at finite @xmath56 .
setting equal to @xmath257 gives the value of the dilaton at the horizon , which can then be applied to to obtain the entropy of a configuration with conical singularity @xmath258 this result takes a more familiar form if we solve for the dilaton as a function of the coordinate @xmath259 @xmath260 the expression for the entropy becomes one - quarter of the horizon area in planck units , even for configurations with a conical singularity @xmath261 the internal energy of these configurations is obtained from . expressed in terms of the boundary conditions and mass parameter @xmath112 , it is @xmath262 for the he s solution @xmath172 the ` hot flat space ' of @xcite the internal energy is zero . for non - zero values of the mass parameter the result can be inverted to give @xmath112 as a function of the internal energy in the region @xmath57 @xmath263 which relates the adm mass to the internal energy @xmath264 and the gravitational binding energy @xmath265 in the cavity @xcite .
in the rest of this section we consider the phase structure of the schwarzschild model , and the dominant contributions to the euclidean partition function coming from black holes with a conical singularity .
it is tempting to focus on the familiar example @xmath266 , but as we will see this is a special case that exhibits qualitatively different behavior than models based on the reduction from @xmath267 dimensions .
the analysis is simpler when quantities are expressed as functions of @xmath114 rather than the mass @xmath112 . in that case
the action is @xmath268 in @xmath266 dimensions this is precisely @xmath67 times the `` generalized free energy '' obtained by york in @xcite .
the action is plotted in figure [ fig : action1 ] as a function of @xmath114 , for representative values of the boundary conditions .
the minimum of the action is either he s or a smooth black hole , depending on the values of @xmath67 and @xmath56 fixed by the boundary conditions .
the he s solution with @xmath213 is always present , but smooth black holes exist in the ensemble only when the smoothness and boundary conditions are both met .
this occurs at isolated values of @xmath215 that satisfy @xmath269 for @xmath270 .
local extrema appear only for @xmath271 below a certain value @xmath272 .
, width=384 ] to analyze this condition it is convenient to define the variables @xmath273 so that takes the form @xmath274 there are no real solutions of , and hence no smooth black holes in the ensemble , for @xmath275 with @xmath276 in this case the action is strictly non - negative and the ground state is he s with @xmath277 .
but if @xmath278 there are two black holes in the ensemble , corresponding to two real solutions @xmath279 of .
the smaller of the two black holes is a local maximum of the action , and the larger black hole is a local minimum .
this minimum is positive when @xmath271 is greater than a critical value given by @xmath280 so that the ground state of the ensemble remains he s for @xmath281 .
but for boundary conditions that satisfy @xmath282 , the minimum of the action is negative and the ground state of the ensemble is the large black hole . using the definition and expressing @xmath56 in terms of the radius of the cavity @xmath283 , the three regimes of the schwarzschild model can be described in terms of more conventional variables . [ cols="^,^,^",options="header " , ] the schwarzschild model has a `` low temperature '' phase , set by the size of the cavity , where smooth black holes do not exist at all black holes in the ensemble necessarily exhibit a conical singularity in this regime .
two smooth black holes appear in the ensemble as the temperature is increased at fixed cavity size .
one of the black holes is stable against small fluctuations ( i.e. , @xmath180 ) , but at intermediate temperatures the system will eventually tunnel from this state to the he s ground state . finally , a `` high temperature '' phase occurs above a critical temperature that is also set by the size of the cavity @xmath284 for @xmath285 the ground state of the ensemble is a smooth black hole .
it is worth taking a moment to consider a gedankenexperiment that examines the phases described above in a `` real world '' setting .
suppose we construct a cavity of macroscopic size in the lab , removing all matter from the interior and holding the walls at a constant temperature .
assuming that gravity is described by the usual einstein - hilbert action ( and neglecting all physics besides gravity and radiation ) , what is the relevant ground state for the ensemble ? restoring dimensionful constants , the condition becomes @xmath286 the two solutions for a cavity of radius @xmath287 held at temperature @xmath288 are @xmath289 the larger solution describes a stable black hole with its horizon about a third of a _ picometer _ from the wall of the cavity ; reasonable laboratory conditions are apparently far into the high - temperature regime !
a quick calculation shows that the stable black hole is the ground state of the ensemble , with a free energy of @xmath290 and an entropy of @xmath291 .
yet the interior of the cavity remains he s , with a free energy of about @xmath292 ( from radiation ) .
what prevents the system from tunneling to the overwhelmingly favorable black hole ground state ?
recall from the analysis of @xcite and @xcite that the rate of tunneling is approximately @xmath293 .
the action of the unstable black hole is enormous , @xmath294 , so the probability of a tunneling event is , for all intents and purposes , zero corresponds to a mass well below the chandrasekhar limit , so the chances of finding one are not good . ] .
contributions to the partition function from field configurations with a conical singularity are approximated to high accuracy by relatively simple functions of the boundary conditions , as described in section [ sec : partitionfunction ] .
for the semiclassical approximation to be valid the action must be very large in units of @xmath295 , which requires @xmath296 in natural units .
in the low and intermediate temperature regimes the ground state is he s , and the contributions to the partition function are approximately @xmath297 this approximation can be compared with a numerical evaluation of .
the fractional error , defined as @xmath298 , is shown in figure [ fig : felowt ] for the case @xmath266 , with @xmath299 and different values of @xmath300 . in the low temperature regime
the error is typically much less than @xmath301 , while in the intermediate temperature regime it is between @xmath301 and @xmath302 for @xmath271 not too close to @xmath303 .
the behavior at the critical temperature is described below . as a function of @xmath304 , for different values of @xmath271 in the low ( @xmath275 ) and intermediate ( @xmath305 ) temperature regimes.,title="fig:",width=388 ] as a function of @xmath304 , for different values of @xmath271 in the low ( @xmath275 ) and intermediate ( @xmath305 ) temperature regimes.,title="fig:",width=384 ] in the high temperature phase the behavior of @xmath102 is qualitatively different .
the approximation for the contributions to the partition function gives @xmath306 the fractional error for this approximation is shown for the @xmath266 model in figure [ fig : fehight ] , with different values of @xmath307 and @xmath308 . as a function of @xmath304 , for different values of @xmath308 in the high temperature regime.,width=384 ] in the high temperature regime @xmath308 takes values in the range @xmath309 which becomes @xmath310 when @xmath266 . for @xmath311 and @xmath312 ,
the error is typically below about @xmath301 .
but at @xmath313 ( when @xmath314 ) the error jumps by 1 - 2 orders of magnitude .
this makes sense ; at the lower end of the smooth black hole has action @xmath315 , so the ground state of the ensemble is a superposition of the black hole and he s . a better approximation for at this transitional value of @xmath308 is given by the sum of and . for the @xmath266 model
this results in a fractional error below @xmath316 .
the dominant contribution to the partition function in the high temperature phase is the overall factor of @xmath317 .
this gives the leading term in the free energy as @xmath318 , and the resulting contribution to the entropy for the smooth black hole is @xmath319 the contributions to @xmath102 from configurations with conical singularities give corrections to @xmath320 and hence to @xmath321 .
the free energy @xmath322 obtained from is @xmath323 which results in an entropy @xmath324 the last two terms combine to give an @xmath325 contribution for all @xmath308 in the range .
thus , the entropy for the @xmath254-dimensional schwarzschild model with corrections from conical singularities takes the form @xmath326 in @xmath266 dimensions there is no @xmath327 correction ; its absence can be traced back to the dependence of various quantities on the boundary condition @xmath56 . in an ensemble that contains smooth black holes , the boundary conditions must satisfy , so for fixed @xmath308 we have @xmath328 .
the specific heat , on the other hand , scales linearly with @xmath56 at fixed @xmath308 @xmath329 the contributions to @xmath330 involve the factor @xmath331 , and for fixed @xmath308 this is independent of @xmath56 when @xmath266 . as a result ,
the corrections to the free energy and the entropy in that case are @xmath325 .
it is important to remember that the canonical partition function for the schwarzschild model is not defined as @xmath93 , despite the fact that some of the results in this section seem well - behaved in that limit . in the next few sections we will consider models that _ do _ admit an @xmath93 limit . in that case
results analogous to simplify quite a bit , and are easier to interpret .
the spherically symmetric reduction of gravity with a negative cosmological constant gives the ads - schwarzschild models .
the functions @xmath332 and @xmath166 in this case are @xmath333 where @xmath334 is the ads length scale and @xmath335 .
since @xmath336 at large @xmath8 , this model satisfies the condition for the existence of the partition function in the @xmath93 limit . rather than considering the ensemble with finite @xmath56 , we will work with ensembles where the cavity wall is removed may be found in @xcite . ] .
following the discussion in section [ sec : partitionfunction ] , the ensemble is defined by fixing the period @xmath241 .
the action for the theory with the cut - off removed is @xmath337 figure [ fig : adsaction ] shows plots of the action for representative values of @xmath241 . the action is extremized by black holes with horizon @xmath142 that satisfy the smoothness condition @xmath338 expressed as a function of @xmath241 , the possible smooth black hole horizons @xmath215 are @xmath339 this presents three different scenarios for smooth black holes in the ensemble , assuming @xmath249 ( the case @xmath340 , the btz black hole , will be discussed separately ) . if @xmath341 then there are no real solutions of , and all black holes in the ensemble have a conical singularity at the horizon .
but if @xmath342 then there are two smooth black holes , with horizons @xmath343 , which correspond to a local maximum ( @xmath344 , with @xmath345 ) and minimum ( @xmath346 , with @xmath347 ) of the action .
the action at the smooth local minimum @xmath346 is @xmath348 the ground state of the ensemble is easily determined by comparing this to the action @xmath226 for the he s solution ( the reduction of ` thermal ads ' ) .
when @xmath349 the action is positive and he s is the ground state of the ensemble ; when @xmath350 the action is negative and ground state is the smooth black hole .
the transition between these phases occurs at a critical value of the period , @xmath351 , which corresponds to a hawking temperature @xmath352 .
thus , there are two phases in the model , which can be divided into three distinct temperature regimes . at low temperatures , @xmath353 ,
there are no smooth black holes in the ensemble and the ground state is he s . for an intermediate range of temperatures , @xmath354 ,
two smooth black holes appear in the ensemble but the ground state remains he s . and finally , at high temperatures , @xmath355 , the ensemble is dominated by the larger of the two smooth black hole solutions .
the approximations derived in section [ sec : partitionfunction ] for the contributions to the partition function should be accurate as long as @xmath356 . in the low temperature regime where he s
dominates the ensemble , @xmath357 , the integral is approximately @xmath358 independent of the dimension @xmath254 of the original model .
the fractional error associated with this approximation is very small ( of order @xmath301 or less for @xmath359 ) and decreases for lower temperatures and larger values of @xmath360 . in the phase of the theory dominated by the smooth black hole , @xmath361
, the integral behaves as @xmath362 where the factor in the exponential is given in . at temperatures less than about twice the critical temperature the fractional error in this approximation can be as large as @xmath363 for @xmath359 , but at higher temperatures or larger values of @xmath360 this rapidly drops to a fraction of a percent . in the high temperature phase the contributions to the partition function are comparable to corrections from ( smooth )
quadratic fluctuations around the ground state .
the free energy is @xmath364 , so @xmath365 where ` @xmath366 ' is a constant . at high temperatures , @xmath367 ,
the relation between the horizon and the temperature is @xmath368 thus , at high temperatures , the entropy of the smooth black hole can be expressed as @xmath369 where @xmath370 is the dominant contribution to the entropy coming from the free energy of the smooth black hole .
the ads model with @xmath340 is an intriguing example where the contributions to the partition function can be computed in terms of elementary functions when @xmath93 .
the condition has a non - zero solution , the btz black hole , with horizon @xmath371 where we have used @xmath372 for @xmath373 .
thus , unlike higher dimensional models , a smooth black hole exists for ensembles with any value of @xmath241 .
the action for the btz black hole is @xmath374 so he s is the ground state of the ensemble for @xmath241 greater than the critical value @xmath375 , and btz is the ground state for @xmath376 .
the action for a black hole with conical singularity in this model is @xmath377 this appears in the integral with the measure @xmath189 , which can be rewritten using the condition @xmath194 to give @xmath378 the integral for contributions to the partition function then takes a form that can be evaluated directly , without the need for approximations @xmath379 evaluating the integral gives @xmath380 in the high temperature regime , @xmath381 , the factor of @xmath382 dominates and @xmath102 is @xmath383 the entropy of the btz is then the usual , leading term , and a sub - leading logarithmic correction with coefficient @xmath384 @xmath385 as with the higher - dimensional ads models , this is precisely the sort of correction that is obtained when smooth fluctuations around the ground state are included in the path integral
. it also agrees with 1-loop calculations , see @xcite and refs . therein .
another example that can be treated in great detail is the jackiw - teitelboim model @xcite , defined by the functions @xmath386 it is convenient to work with the location of the horizon , rather than the mass parameter @xmath112 . then the metric function for a black hole with conical singularity is @xmath387 , and the action for such
a configuration is @xmath388 the smoothness condition that extremizes the action yields @xmath389 inverting this expression for @xmath215 identifies a single smooth black hole that is present in the ensemble for all values of @xmath67 and @xmath56 @xmath390 with action @xmath391 this expression is always negative , which means that the smooth black hole dominates he s ( with action @xmath392 ) . unlike the previous examples ,
the conical ensemble for the jackiw - teitelboim model always has a smooth black hole for the ground state .
the function @xmath15 for the jackiw - teitelboim model satisfies the condition , which means that the @xmath93 limit of the ensemble exists . before considering the contributions to the partition function for the ensemble with finite @xmath56 , let us examine this simpler case .
the action for the ensemble with the cavity wall removed is @xmath393 which is minimized by a smooth black hole with @xmath394 .
the action at this minimum is @xmath395 , and the contributions to the partition function are @xmath396 as with the btz black hole , this integral can be evaluated in closed form to give @xmath397 for the semiclassical approximation to hold , the action of the minimum should be large in units of @xmath295 , which requires @xmath398 in natural units . in that case
the entropy obtained from is @xmath399 the first term in this expression is the leading contribution to the entropy of the smooth black hole , and the second term represents a correction due to the contributions from conical singularities .
the correction once again coincides with the result when smooth quadratic fluctuations around the regular ground state are taken into account @xcite . for finite @xmath56 the contributions to the semiclassical partition function
can not be evaluated exactly , but they are approximated to a high degree of accuracy by a relatively simple function of the boundary conditions .
we find @xmath400 where @xmath401 is the action for the smooth black hole .
the fractional error for this approximation , compared to a numerical evaluation of , is shown in figure [ fig : jtfe ] . , as a function of @xmath304 , for different ratios of horizon to cavity size .
each curve is labeled by the value of @xmath402.,width=432 ] the entropy calculated using consists of a leading term and corrections @xmath403 where ` @xmath366 ' indicates terms that are independent of @xmath404 .
the first term in the entropy is the leading behavior associated with the smooth black hole , and the next two terms are the corrections from conical singularity contributions . for small values of @xmath61
is approximately @xmath405 which exhibits the same form @xmath406 as the ensemble with the cavity wall removed .
this is not too surprising , since @xmath407 must be held fixed in the @xmath93 limit , which implies @xmath408 . in this section
we consider three models related to black holes that arise in string theory , either as solutions of the @xmath409-functions at lowest order in @xmath410 , or as exact solutions that incorporate corrections at all orders in @xmath410 .
there is an important difference between this section and the previous ones : the stringy models must be considered with the cavity wall removed . in
string theory one can not introduce new degrees of freedom in an arbitrary manner ( as we have done with the thermal reservoir ) , and _ ad hoc _ cut - offs on spacetime fields ( the restriction @xmath57 on the dilaton ) are usually equivalent to a truncation of states on the worldsheet that spoil the consistency of the theory .
since the models we consider all have non - compact target spaces with @xmath55 asymptotically , the @xmath93 limit is required for any calculations that are meant to be interpreted in the context of string theory .
additional discussion can be found in @xcite .
it is interesting that there is an intrinsic way to identify string - like behavior within the zoo of two - dimensional dilaton gravity models .
namely , a universal property of all stringy models is that the weyl - invariant function @xmath166 is linear in the dilaton field @xmath8 for large values of the dilaton ( which means weak coupling from a target - space perspective ) .
linearity of @xmath166 is not just a technical curiosity , but has important physical implications : by virtue of the inequality stringy models always exhibit a hagedorn temperature .
the existence of a hagedorn temperature ( or , equivalently , asymptotic linearity of @xmath166 ) therefore can be considered as a defining property of stringy models .
the witten black hole @xcite is obtained from a solution of bosonic string theory with the worldsheet dynamics described by a @xmath411 coset model .
when the level of the worldsheet current algebra is taken to be large , the tree - level @xmath409-functions at lowest order in @xmath410 have a black hole solution of the form .
the equations may be obtained from an action with @xmath10 and @xmath11 such that @xmath412 with the positive parameter @xmath413 related to the string scale as @xmath414 .
the condition implies that the @xmath93 limit is only defined for this model if @xmath415 . in this
limit the action is @xmath416 the smoothness condition that extremizes the action gives a single value of @xmath241 for which a smooth black hole exists : @xmath417 this corresponds to the hagedorn temperature for the model , at which point contributions from states with a conical singularity at the horizon cause the partition function to diverge .
therefore the @xmath93 limit of this model does not admit an ensemble containing a smooth black hole .
however , one can see from that black hole solutions would have scalar curvature of order @xmath418 near the horizon , indicating that @xmath410 corrections are important . in the next section we will consider a model that takes these corrections into account and always has a black hole ground state . before moving on it is worth considering two more aspects of the model .
first , the partition function for this model is well - defined in the @xmath93 limit as long as @xmath419 , in which case the ground state is he s .
the integral can be directly evaluated , and gives @xmath420 where @xmath421 is the hagedorn temperature .
second , despite problems with implementing a finite @xmath56 cut - off in string theory , one might consider this model as an example of a dilaton gravity where the contributions to the partition function can be calculated exactly for finite @xmath56 .
the action is @xmath422 with local extrema @xmath215 given by the smoothness condition @xmath423 the ensemble contains a single smooth black hole if @xmath424 , and the action for this configuration is always negative .
thus , there are two phases : a he s ground state for @xmath425 , and a smooth black hole ground state for @xmath424 . the contributions to the partition function from conical singularities can be calculated exactly in either phase by expressing the action in terms of @xmath67 and @xmath426 @xmath427 then the integral is given in terms of exponentials and error functions .
the exact form is not especially enlightening , but the behavior simplifies when @xmath428 . in the low temperature phase @xmath425
we obtain essentially the same result as @xmath429 at high temperatures , @xmath430 , the smooth black hole dominates and the contributions to are approximately @xmath431 where the factor in the exponential is minus the action for the smooth black hole .
the free energy is @xmath432 , and the resulting entropy is @xmath433 unlike the previous examples , the correction does not appear to be proportional to @xmath327
. however , the approximate result assumes @xmath428 , and this assumption must be treated carefully since the model does not have an @xmath93 limit above the hagedorn temperature @xmath434 .
the large-@xmath56 result can only be trusted if @xmath67 remains much less than @xmath435 , which implies that @xmath402 must be close to @xmath436 .
then becomes @xmath437 this is , in a sense , the expected result : the functions that define this model may also be thought of as the @xmath438 limit of the schwarzschild model , and is indeed the @xmath438 limit of .
note , though , that the inverse specific heat for the schwarzschild model goes to zero in this limit , so perhaps a better interpretation of is the leading term in a @xmath439 expansion for large but finite @xmath440 .
the black hole background with @xmath410 corrections taken into account was studied in @xcite .
the worldsheet theory is described by an @xmath411 gauged wzw model with level @xmath441 . as shown in @xcite , the exact string black hole corresponds to a dilaton gravity model with @xmath442 where the field @xmath443 is related to the conventionally defined dilaton @xmath8 by @xmath444 the parameter @xmath445 depends on both the level of the model and the string tension @xmath446 for a critical string theory with a target space of dimension @xmath447 , it satisfies the condition @xmath448 normally this would fix @xmath449 at a specific value ( @xmath450 for a critical string theory in two dimensions ) , but as in @xcite we will assume that extra matter fields are present that contribute to the central charge .
this modifies the condition , which has the effect of allowing us to consider other values of @xmath449 . in practice ,
@xmath445 is treated as a fixed parameter and the level takes values in the range @xmath451 .
the value of the field @xmath443 at the horizon is related to the level of the cft by @xmath452 the smoothness condition with the cavity wall removed has solutions for any value of @xmath441 @xmath453 corresponding to a black hole of mass @xmath454 as expected , this black hole is always the ground state of the theory .
the on - shell action for the black hole is explicitly negative for all @xmath441 @xmath455 where we have restored factors of the two - dimensional newton s constant @xmath456 .
computing the entropy from the leading term in the free energy @xmath457 gives @xmath458 now , according to the approximation in section [ sec : partitionfunction ] , the contributions to the partition function from configurations with a conical singularity give @xmath459 which suggests a sub - leading correction to the free energy of the form @xmath460 and consequently a correction to the entropy of the form @xmath461 since we are considering a solution of the genus - zero @xmath409-functions , we must take @xmath462 to ensure that the string coupling is small for any value of @xmath449 . then , in the semiclassical limit @xmath463 the results for @xmath464 and @xmath247 simplify @xmath465 we arrive at an interesting result ; the corrected entropy takes the form @xmath466 the genus - zero beta functions of type 0a string theory have a solution at leading order in @xmath410 that describes a two - dimensional black hole with constant ramond - ramond flux @xcite . this can be viewed as a solution of a model with functions @xmath166 and @xmath332 given by @xmath467 where @xmath468 is positive , and @xmath469 is proportional to the flux of each of the two ramond - ramond gauge fields , in the conventions of @xcite .
this is related to the parameter in @xmath15 by @xmath470 . ] .
the ratio @xmath471 approaches @xmath413 in the limit @xmath55 , indicating that the ensemble is only defined at temperatures below @xmath472 when the cavity wall is removed .
this is the same hagedorn temperature as the witten black hole model , but unlike that model the ensemble now contains a smooth black hole .
the smoothness condition gives @xmath473 which identifies a single black hole @xmath42 for any @xmath474 below the hagedorn temperature @xmath475 combined with the upper limit on the temperature , this result implies that the dilaton at the horizon satisfies @xmath476 .
indeed , it turns out that @xmath477 sets a lower bound on the dilaton at the horizon for any configuration in the ensemble , with or without a conical singularity .
this is due to the fact that @xmath15 has a minimum at @xmath478 ; the requirement @xmath479 for @xmath480 then implies @xmath481 .
the action for the smooth black hole is @xcite @xmath482 and the action for a configuration with a conical singularity is @xmath483 to determine the ground state , we must compare the action for the smooth black hole to the action for the configuration with @xmath114 taking the the minimum value @xmath484 .
their difference is @xmath485 which is always negative , since the ratio @xmath486 is less than one .
thus , the ground state of the ensemble is always the smooth black hole .
given the action for the smooth black hole , the leading contribution to the free energy is @xmath487 .
the resulting entropy is @xmath488 we may now consider corrections from configurations with conical singularities .
the integral can be evaluated exactly using incomplete gamma functions , but for our purposes the approximation is sufficient @xmath489 the contribution to the free energy is @xmath490 and the contribution to the entropy is @xmath491 as in the previous two examples , the relation between @xmath247 and @xmath464 is not clear until we consider the conditions for the semiclassical approximation . with
the cavity wall removed , the semiclassical limit is obtained by taking @xmath492 . in that case @xmath247 becomes @xmath493 which is the same general form as the witten model at finite @xmath56 .
we have considered black holes in an unsuitable box a cavity coupled to a thermal reservoir at a temperature that is in general different than the hawking temperature and studied the thermodynamics of a ` conical ensemble ' that includes these spaces alongside the more conventional smooth field configurations .
the focus was on black holes that allow an effective description in terms of 2-dimensional dilaton gravity , including schwarzschild , schwarzschild - ads , jackiw - teitelboim and various stringy black holes .
we demonstrated that smooth solutions of the equations of motion are locally ( perturbatively ) stable against singular configurations with a small angular deficit or surplus , proved that the ground state of the conical ensemble never exhibits a conical singularity , and calculated corrections to the entropy and free energy for several pertinent examples . in all the examples we considered
, configurations with a conical singularity result in corrections to the entropy that take the same form as generic logarithmic corrections from thermal ( and in some cases also quantum ) fluctuations .
in fact , our results can be compared with previous results for entropy corrections from these sorts of fluctuations if the following caveats are taken into account : 1 . in many cases
existing results have been obtained in the microcanonical ensemble .
translating our results , derived in the canonical ensemble , into corrections for the microcanonical entropy changes the sign of the coefficient of the logarithmic term .
matter interactions and non - spherical excitations have been neglected , so naturally we can compare only with results where these contributions are switched off .
3 . in the schwarzschild model
the partition function does not exist in the @xmath93 limit , so our results are only meaningful for a finite cavity . with these caveats in mind , we can present the schwarzschild result also as a correction to the microcanonical entropy ( which coincides at leading order with the canonical entropy @xmath464 ) @xmath494 here @xmath495 refers to all matter fields and graviton excitations ( basically their contributions to the trace anomaly ) and @xmath496 is a separate contribution from @xmath497 gauge fields ; our simple approach is not sensitive to either of these contributions .
the result agrees precisely with eq .
( 1.4 ) in @xcite when matter fields are switched off and the graviton excitations are frozen ) contributions .
therefore , this is the appropriate ensemble to compare with .
the 4-dimensional microcanonical result contains an additional contribution coming from the cartan generators of the rotation group , to which our analysis naturally is blind . ] .
likewise , the microcanonical analogs of the entropy corrections for the ads - schwarzschild model and for the btz black hole are the same as the results obtained in @xcite .
we consider this to be a consequence of the semiclassical approximation , where the leading corrections to the partition function are given by a gaussian integral . in both cases
conical singularities and thermal fluctuations
the coefficient of the quadratic term in the exponent is proportional to @xmath498 ( or @xmath499 ) , leading to similar corrections . in our analysis
we have only taken into account configurations with a single conical defect , located at the black hole horizon .
a possible generalization is the inclusion of multiple conical defects , which are not necessarily located at the horizon .
this is challenging for at least two reasons .
first , the existence and description of multiple conical singularities on a given space is an open problem that depends on curvature bounds and other model - specific quantities @xcite .
second , the action is not suitable for this purpose .
we have assumed that field configurations in the ensemble exhibit the same symmetries as the cavity . from the point of view of a higher dimensional theory ,
the cavity is spherical and elements of the ensemble are spherically symmetric . including less symmetric configurations in the ensemble requires a more general action that contains terms not present in , and ignoring the contributions from these terms leads to nonsensical results .
for instance , it is tempting to try and study a conical singularity somewhere between the horizon of a smooth black hole and the cavity wall by replacing @xmath500 in with @xmath501 , for some @xmath502 . in any model that admits an @xmath93 limit
, this gives an action that is unbounded below for @xmath503 .
but this pathological behavior is simply the result of neglecting important contributions to the action ; there is no catastrophic instability that suddenly produces conical singularities at large @xmath504 . one obvious shortcoming of our analysis is that it applies only to black holes that are symmetric enough to allow an effective 2-dimensional description in terms of a dilaton gravity model with action .
it would be of interest to lift our results to higher dimensions , particularly to three and four dimensions . as an example of what one could learn from such a generalization
let us focus on three dimensions .
a few years ago the interest in 3-dimensional ( quantum ) gravity was rekindled , see e.g. @xcite .
in particular , maloney and witten showed that the euclidean partition function of pure einstein gravity with a negative cosmological constant is not a sensible cft partition function and does not factorize holomorphically @xcite .
they arrived at their result by taking into account all known contributions to the euclidean partition function on the gravity side , assuming smooth metrics , and speculated ( among other logical possibilities ) that the partition function could be made sensible by taking into account configurations with a conical defect .
given the results of the present work this option does not seem to be likely : we have demonstrated in all explicit examples that the leading contribution from conical defects to the partition function behave in the same way as the leading contributions from thermal or quantum fluctuations .
if it remains true in the presence of conical defects that the partition function of 3-dimensional einstein gravity is 1-loop exact , this means that the qualitative features of the partition function are unlikely to change dramatically upon inclusion of conical defects .
it would be of interest to demonstrate this explicitly by lifting our results to three dimensions .
we thank roman jackiw , rob myers and rafael sorkin for useful discussions and feedback on an early version of this work .
dg is supported by the start project y435-n16 of the austrian science fund ( fwf ) .
rm is supported by a faculty research stipend from loyola university chicago .
sz benefits from a phd research grant of the institut interuniversitaire des sciences nuclaires ( iisn , belgium ) ; his work is supported by the belgian federal office for scientific , technical and cultural affairs through the interuniversity attraction pole p6/11 .
dg and rm thank the perimeter institute and the center for theoretical physics at mit for hospitality and support during the early stages of this work .
finally , rm would like to acknowledge the birth of his wonderful daughter willa .
her arrival in february 2012 provided him with his best excuse yet for not completing a project on schedule .
s. carlip and c. teitelboim , `` the off - shell black hole , '' _ class.quant.grav . _
* 12 * ( 1995 ) 16991704 , http://www.arxiv.org/abs/gr-qc/9312002[gr-qc/9312002 ] .
d. j. gross , m. j. perry , and l. g. yaffe , `` instability of flat space at finite temperature , '' _ phys .
d25 * ( 1982 ) 330355 . s. w. hawking and d. n. page , `` thermodynamics of black holes in anti - de sitter space , '' _ commun .
* 87 * ( 1983 ) 577 .
s. das , p. majumdar , and r. k. bhaduri , `` general logarithmic corrections to black hole entropy , '' _ class .
* 19 * ( 2002 ) 23552368 , http://www.arxiv.org/abs/hep-th/0111001[hep-th/0111001 ] .
s. n. solodukhin , `` the conical singularity and quantum corrections to entropy of black hole , '' _ phys.rev . _
* d51 * ( 1995 ) 609617 , http://www.arxiv.org/abs/hep-th/9407001[hep-th/9407001 ] .
a. sen , `` logarithmic corrections to schwarzschild and other non - extremal black hole entropy in different dimensions , '' http://www.arxiv.org/abs/1205.0971[1205.0971 ] .
d. grumiller and r. mcnees , `` thermodynamics of black holes in two ( and higher ) dimensions , '' _ jhep _ * 04 * ( 2007 ) 074 , http://www.arxiv.org/abs/hep-th/0703230[hep-th/0703230 ] .
d. grumiller , w. kummer , and d. v. vassilevich , `` dilaton gravity in two dimensions , '' _ phys .
_ * 369 * ( 2002 ) 327429 , http://arxiv.org/abs/hep-th/0204253[hep-th/0204253 ] .
d. grumiller and r. meyer , `` ramifications of lineland , '' _ turk . j. phys .
_ * 30 * ( 2006 ) 349378 , http://www.arxiv.org/abs/hep-th/0604049[hep-th/0604049 ]
. j. w. york , jr .
, `` role of conformal three geometry in the dynamics of gravitation , '' _ phys .
_ * 28 * ( 1972 ) 10821085 .
g. w. gibbons and s. w. hawking , `` action integrals and partition functions in quantum gravity , '' _ phys .
_ * d15 * ( 1977 ) 27522756 .
d. grumiller , `` path integral for half - binding potentials as quantum mechanical analog for black hole partition functions , '' http://www.arxiv.org/abs/0711.4115[0711.4115 ] .
d. grumiller and p. van nieuwenhuizen , `` holographic counterterms from local supersymmetry without boundary conditions , '' _ phys.lett .
_ * b682 * ( 2010 ) 462465 , http://www.arxiv.org/abs/0908.3486[0908.3486 ] .
schmidt , `` scale invariant gravity in two - dimensions , '' _ j. math
* 32 * ( 1991 ) 15621566 .
t. banks and m. oloughlin , `` two - dimensional quantum gravity in minkowski space , '' _ nucl .
* b362 * ( 1991 ) 649664 .
j. gegenberg , g. kunstatter , and d. louis - martinez , `` observables for two - dimensional black holes , '' _ phys .
_ * d51 * ( 1995 ) 17811786 , http://www.arxiv.org/abs/gr-qc/9408015[gr-qc/9408015 ] .
r. c. tolman , _ relativity , thermodynamics , and cosmology_. oxford university press , oxford , 1934 .
j. w. york , jr .
, `` black hole thermodynamics and the euclidean einstein action , '' _ phys . rev . _
* d33 * ( 1986 ) 20922099 .
g. w. gibbons and m. j. perry , `` the physics of 2-d stringy space - times , '' _ int .
* d1 * ( 1992 ) 335354 , http://arxiv.org/abs/hep-th/9204090[hep-th/9204090 ] .
e. farhi , a. h. guth , and j. guven , `` is it possible to create a universe in the laboratory by quantum tunneling ?
, '' _ nucl .
* b339 * ( 1990 ) 417490 .
g. hayward , `` gravitational action for space - times with nonsmooth boundaries , '' _ phys .
_ * d47 * ( 1993 ) 32753280 .
d. brill and g. hayward , `` is the gravitational action additive ?
, '' _ phys .
* d50 * ( 1994 ) 49144919 , http://www.arxiv.org/abs/gr-qc/9403018[gr-qc/9403018 ] .
d. v. fursaev and s. n. solodukhin , `` on the description of the riemannian geometry in the presence of conical defects , '' _ phys .
rev . _ * d52 * ( 1995 ) 21332143 , http://www.arxiv.org/abs/hep-th/9501127[hep-th/9501127 ] .
l. landau and e. lifshitz , _ course of theoretical physics _ , vol .
5 : statistical physics . london , 1959 .
m. o. katanaev , w. kummer , and h. liebl , `` on the completeness of the black hole singularity in 2d dilaton theories , '' _ nucl .
phys . _ * b486 * ( 1997 ) 353370 , http://www.arxiv.org/abs/gr-qc/9602040[gr-qc/9602040 ] .
m. m. akbar , `` schwarzschild - anti de sitter within an isothermal cavity : thermodynamics , phase transitions and the dirichlet problem , '' http://www.arxiv.org/abs/hep-th/0401228[hep-th/0401228 ] .
r. jackiw , `` liouville field theory : a two - dimensional model for gravity ?
, '' in _ quantum theory of gravity _ , s. christensen , ed .
. 403420 .
adam hilger , bristol , 1984 . c. teitelboim , `` the hamiltonian structure of two - dimensional space - time and its relation with the conformal anomaly , '' in _ quantum theory of gravity _ , s. christensen , ed . , pp .
adam hilger , bristol , 1984 .
d. grumiller , `` logarithmic corrections to the entropy of the exact string black hole , '' in _ path integrals from quantum information to cosmology _ , c. burdik ,
n. navratil , and s. posta , eds .
jinr publishing department , prague , june , 2005 .
http://www.arxiv.org/abs/hep-th/0506175[hep-th/0506175 ] .
e. witten , `` on string theory and black holes , '' _ phys .
rev . _ * d44 * ( 1991 ) 314324 . c. r. nappi and a. pasquinucci ,
`` thermodynamics of two - dimensional black holes , '' _ mod . phys .
lett . _ * a7 * ( 1992 ) 33373346 , http://arxiv.org/abs/gr-qc/9208002[gr-qc/9208002 ] .
r. dijkgraaf , h. verlinde , and e. verlinde , `` string propagation in a black hole geometry , '' _ nucl .
phys . _ * b371 * ( 1992 ) 269314 .
d. grumiller , `` an action for the exact string black hole , '' _ jhep _ * 05 * ( 2005 ) 028 , http://www.arxiv.org/abs/hep-th/0501208[hep-th/0501208 ] .
v. a. kazakov and a. a. tseytlin , `` on free energy of 2-d black hole in bosonic string theory , '' _ jhep _ * 06 * ( 2001 ) 021 , http://arxiv.org/abs/hep-th/0104138[hep-th/0104138 ] .
n. berkovits , s. gukov , and b. c. vallilo , `` superstrings in 2d backgrounds with r - r flux and new extremal black holes , '' _ nucl .
* b614 * ( 2001 ) 195232 , http://arxiv.org/abs/hep-th/0107140[hep-th/0107140 ] .
m. r. douglas _ et al .
_ , `` a new hat for the c = 1 matrix model , '' http://www.arxiv.org/abs/hep-th/0307195[hep-th/0307195 ]
. j. l. davis and r. mcnees , `` boundary counterterms and the thermodynamics of 2-d black holes , '' _ jhep _ * 09 * ( 2005 ) 072 , http://www.arxiv.org/abs/hep-th/0411121[hep-th/0411121 ] .
m. troyanov , `` prescribing curvature on compact surfaces with conical singularities , '' _ trans .
* 324 * ( 1991 ) , no . 2 , 793821 . e. witten , `` three - dimensional gravity revisited , '' http://www.arxiv.org/abs/0706.3359[0706.3359 ] .
w. li , w. song , and a. strominger , `` chiral gravity in three dimensions , '' _ jhep _ * 04 * ( 2008 ) 082 , http://www.arxiv.org/abs/0801.4566[0801.4566 ] .
s. carlip , s. deser , a. waldron , and d. k. wise , `` cosmological topologically massive gravitons and photons , '' _ class .
* 26 * ( 2009 ) 075008 , http://www.arxiv.org/abs/0803.3998[0803.3998 ] .
d. grumiller and n. johansson , `` instability in cosmological topologically massive gravity at the chiral point , '' _ jhep _ * 07 * ( 2008 ) 134 , http://www.arxiv.org/abs/0805.2610[0805.2610 ] .
a. maloney and e. witten , `` quantum gravity partition functions in three dimensions , '' _ jhep _ * 1002 * ( 2010 ) 029 , http://www.arxiv.org/abs/0712.0155[0712.0155 ] . | we consider black holes in an `` unsuitable box '' : a finite cavity coupled to a thermal reservoir at a temperature different than the black hole s hawking temperature .
these black holes are described by metrics that are continuous but not differentiable due to a conical singularity at the horizon .
we include them in the euclidean path integral sum over configurations , and analyze the effect this has on black hole thermodynamics in the canonical ensemble .
black holes with a small deficit ( or surplus ) angle may have a smaller internal energy or larger density of states than the nearby smooth black hole , but they always have a larger free energy .
furthermore , we find that the ground state of the ensemble never possesses a conical singularity .
when the ground state is a black hole , the contributions to the canonical partition function from configurations with a conical singularity are comparable to the contributions from smooth fluctuations of the fields around the black hole background .
our focus is on highly symmetric black holes that can be treated as solutions of two - dimensional dilaton gravity models : examples include schwarzschild , asymptotically anti - de sitter , and stringy black holes . |
in this paper we are interested in knots residing in spatial embeddings of graphs . before describing our interests
concretely we need to give necessary definitions . a circle embedded into the euclidean 3-space @xmath1
is called a _
knot_. two knots @xmath2 and @xmath3 are said to be _ ambient isotopic _ , denoted by @xmath4 , if there exists a continuous map
@xmath5 \rightarrow \mathbb{r}^3 $ ] such that the restriction of @xmath6 to each @xmath7 $ ] , @xmath8 , is a homeomorphism , @xmath9 is the identity map and @xmath10 , to say roughly , @xmath11 can be deformed to @xmath12 without intersecting its strand . the ambient isotopy class of a knot @xmath2
is called the _ knot type _ of @xmath2 .
especially if @xmath2 is ambient isotopic to another knot contained in a plane of @xmath1 , then we say that @xmath2 is _
trivial_. a _ polygonal knot _ is a knot consisting of finitely many line segments .
figure [ fig1 ] shows polygonal presentations of two knot types @xmath13 and @xmath14 ( these notations for knot types follow the knot tabulation in @xcite .
usually @xmath13 and its mirror image are called _ trefoil _ , and @xmath14 _ figure-8 _ ) .
polygonal knots appear in many scientific contexts other than mathematics .
real knots such as polymers are modeled to be spatial polygons for theoretical studies on their physical and chemical properties @xcite .
one quantity of polygonal knots interesting in the related research is polygonal index . for a knot
type @xmath15 , its _ polygonal index _
@xmath16 is defined to be the minimal number of line segments required to realize @xmath15 as a polygon . from the definition we easily know @xmath17 .
but generally it is not easy to determine @xmath16 for an arbitrary knot type @xmath15 .
this quantity was determined only for some specific knot types @xcite . here
we mention a result by randell on small knots for later use .
[ stick - thm ] @xcite @xmath17 , @xmath18 and @xmath19 .
furthermore , @xmath20 for any other knot type @xmath15 .
let @xmath21 be the complete graph with @xmath22 vertices .
a spatial embedding of @xmath21 , that is , an embedding of @xmath21 into @xmath1 , will be said to be _ linear _ , if each edge of the graph is mapped to a line segment .
note that a linearly embedded @xmath21 contains @xmath23 polygonal knots as its hamiltonian cycles . in 1983
conway and gordon proved that any spatial embedding of @xmath0 contains at least one nontrivial knot as its hamiltonian cycle @xcite .
and their result was generalized by negami .
he showed that for any knot type @xmath15 , there exists a natural number @xmath24 such that every linearly embedded @xmath25 contains a polygonal knot of the type @xmath15 @xcite .
for example @xmath26 @xcite . after conway - gordon s work , knots ( also links )
are considered as intrinsic properties of abstract graphs , and numerous subsequent works have been continued until recently . for a survey on this research the readers are referred to @xcite . in this paper
we would like to study knots in @xmath0 more specifically .
let @xmath27 be a linear embedding . from theorem [ stick - thm ]
we know that any spatial trigon , tetragon and pentagon are trivial knots .
hence , for @xmath28 and @xmath29 , @xmath30 contains only trivial knots as its cycles . by theorem [ stick - thm ]
any spatial hexagon is trivial or trefoil , and trefoil is the only nontrivial knot type which may appear in @xmath31 .
hence we may ask how many hexagonal trefoil knots can reside in @xmath31 as hamiltonial cycles .
this question was answered in a previous work of the author @xcite , in which the point configuration of vertices of hexagonal trefoil knot was investigated and it was proved that _ @xmath31 contains at most one trefoil knot_. recently , using the second coefficient of the conway polynomial for knots , this was reproved by nikkuni @xcite . again by theorem [ stick - thm ] any heptagon is trivial , trefoil or figure-8 , and hence the figure-8 is the knot type which may be newly found in @xmath32 . and
, considering polygonal index to be a measure for knot complexity , we may say that the figure-8 is the largest knot type which can reside in @xmath32 . in this paper
we determine how many heptagonal figure-8 knots can be contained in @xmath32 .
the following is the main result of this paper .
[ main - thm ] any linear embedding of @xmath0 contains at most three heptagonal figure-8 knots as its hamiltonian cycles .
note that a generic set of seven points in @xmath1 determine a linear embedding of @xmath0 .
hence the theorem can be restated as follows : any seven points in general position of @xmath1 constitute at most three heptagonal figure-8 knots .
figure [ fig0 ] illustrates a realization of linearly embedded @xmath0 which contains three figure-8 knots @xmath33 , @xmath34 and @xmath35 .
therefore the upper bound in theorem [ main - thm ] is strict .
the rest of paper will be devoted to the proof of theorem [ main - thm ] .
in this section we introduce a previous result of the author @xcite which would be a key lemma to prove theorem [ main - thm ] .
let @xmath36 be a heptagonal knot .
we will call the line segments of @xmath36 _ edges of @xmath36 _ and their end points _ vertices_. the vertices of @xmath36 are assumed to be in general position , that is , any four vertices are not coplanar . and label the vertices of @xmath36 by @xmath37 so that each vertex @xmath38 is connected to @xmath39 ( mod @xmath40 ) by an edge of @xmath36 , that is , a labeling of vertices is determined by a choice of base vertex and an orientation of @xmath36 . given a labelling of vertices let @xmath41 denote the triangle formed by three vertices @xmath42 , and @xmath43 the line segment from the vertex @xmath44 to @xmath45 .
the relative position of such a triangle and a line segment will be represented via `` @xmath46 '' which is defined in the below : * if @xmath47 , then set @xmath48 . * otherwise , + @xmath49 ( _ resp . @xmath50 _ ) , when @xmath51 ( _ resp .
@xmath52 _ ) .
the tables in theorem [ key - lemma ] show the values of @xmath46 between triangles formed by three consecutive vertices and edges of @xmath36 .
if @xmath46 is zero , then the corresponding cell in the table is filled by `` @xmath53 '' .
otherwise , we mark by `` @xmath54 '' or `` @xmath55 '' according to the sign of @xmath46 . for example , if @xmath36 is of type - i , then @xmath56 and + @xmath57 .
@xcite [ key - lemma ] let @xmath36 be a heptagonal knot such that its vertices are in general position
. then @xmath36 is figure-8 if and only if the vertices of @xmath36 can be labelled so that the polygon satisfies one among three types type - i , ii and iii .
let @xmath2 be a linearly embedded @xmath0 which contains a heptagonal figure-8 knot @xmath36 as its hamiltonian cycle .
theorem [ key - lemma ] guarantees that the vertices of @xmath2 can be labelled by @xmath58 so that @xmath36 corresponds to the cycle @xmath59 and satisfies one among type - i , ii and iii .
therefore our main theorem comes from the three lemmas in the below .
[ lemma-1 ] if @xmath36 is of type - i , then @xmath2 contains at most two heptagonal figure-8 knots as its hamiltonian cycles .
[ lemma-2 ] if @xmath36 is of type - ii , then @xmath2 contains at most three heptagonal figure-8 knots as its hamiltonian cycles .
[ lemma-3 ] if @xmath36 is of type - iii , then it is the only figure-8 knot among the hamiltonian cycles of @xmath2 .
these lemmas will be proved in next three sections . to prove them we utilize a concept , called _
trivial triple_. a set of three vertices @xmath60 of @xmath2 will be called a trivial triple , if @xmath61 for every two vertices @xmath62 .
suppose that a heptagon @xmath63 in @xmath2 has the trivial triple as its consecutive three vertices .
then @xmath63 is not figure-8 , because we can isotope @xmath63 to be a hexagon by reduction along @xmath64 as illustrated in figure [ fig2 ] . in each case of type - i ,
ii and iii we will try to find trivial triples as many as possible , so that heptagons which have any of such triples as consecutive vertices are excluded from possible hamiltonian cycles of figure-8 . and then the remaining hamiltonian cycles will be observed more closely to check whether they are figure-8 or not .
throughout this section we suppose that @xmath36 is of type - i .
then , without loss of generality , it may be assumed that @xmath65 .
we begin the proof of lemma [ lemma-1 ] with three claims in the below .
let @xmath66 denote the plane containing three vertices @xmath60 .
for an ordered sequence @xmath67 of the three vertices , define @xmath68 * claim ( i ) . *
_ the plane @xmath69 separates the two vertices @xmath70 from @xmath71 . _ + _ proof .
_ under our assumption , @xmath72 is @xmath50 , hence we know that @xmath69 separates the vertex @xmath73 from @xmath74 as illustrated in figure [ fig3 ] .
in fact the vertex @xmath73 ( _ resp . _ 3 ) belongs to @xmath75 ( _ resp .
_ @xmath76 ) . also from @xmath77
we have that @xmath78 and @xmath79 .
@xmath80 from claim ( i ) we can find out some trivial triples .
for example the interior of @xmath81 belongs to @xmath76 , but the four vertices @xmath82 do not belong to the open half space .
this implies that there is no edge of @xmath2 which penetrates @xmath81 . in this way we have a set of trivial triples in the below . @xmath83
* claim ( ii ) . * _ the plane @xmath84 separates the two vertices @xmath85 from @xmath86 . _
+ * claim ( iii ) . * _ the plane @xmath87 separates the two vertices @xmath88 from @xmath89 .
_ + claim ( ii ) can be proved from @xmath77 and @xmath90 in a similar way .
also a set of trivial triples is derived from the claim .
let @xmath94 be a plane orthogonal to @xmath95 and @xmath96 be the orthogonal projection onto @xmath94 such that the vertex @xmath74 is above the vertex @xmath73 with respect to the @xmath97-coordinate .
then @xmath98 is projected to a line segment .
now we observe the projected image of @xmath36 under @xmath99 . since @xmath92 and @xmath100 , the edges @xmath101 and @xmath102 should pass over @xmath103 with respect to the @xmath97-coordinate .
also @xmath104 implies that @xmath105 passes over @xmath102 .
figure [ fig4]-(a ) shows the image of edges from @xmath106 to @xmath107 under @xmath99 .
note that @xmath108 should be located in the shaded region , because the edge @xmath109 penetrates @xmath110 .
now we observe the relative position of @xmath111 with respect to @xmath101 and @xmath102 .
@xmath112 belongs to the outside of @xmath113 , and @xmath111 penetrates @xmath114 .
this implies that @xmath115 should intersect both @xmath116 and @xmath117 . on
the other hand @xmath101 penetrates @xmath118 , and as shown in figure [ fig4]-(a ) the edge passes over @xmath103 . hence @xmath101 should pass under @xmath111 . consequently , for @xmath111 to penetrate @xmath114 , it should pass under @xmath102 . from the resulting image of @xmath119 in figure [ fig4]-(b ) we immediately find three more trivial triples @xmath120 also from the image we know that @xmath121 is the only edge of @xmath2 which may penetrate @xmath122 .
but this is not possible because @xmath123 , that is , @xmath124 separates the vertex @xmath125 from @xmath29 .
hence @xmath126 is another trivial triple . from claim ( i ) we know that @xmath127 , @xmath128 and @xmath129 are the only edges of @xmath2 that may penetrate @xmath130 .
but again the image of @xmath36 enables us to exclude @xmath128 and @xmath129 from the possibility .
also the edge @xmath127 should be excluded , because @xmath101 penetrates @xmath118 , that is , @xmath131 belongs to @xmath118 .
therefore we have the fifth set of trivial triples @xmath132 to find more trivial triples we consider two subcases according to the position of the vertex @xmath40 with respect to the plane @xmath133 .
figure [ fig4]-(b ) shows that the vertices @xmath134 belong to the half space @xmath135 .
if the vertex @xmath40 belongs to the other half space @xmath136 , then @xmath137 is a trivial triple .
now suppose that the vertex belongs to @xmath135 .
then @xmath138 is trivial , because the vertex @xmath107 is the only vertex in @xmath136 .
we will show that two more triples @xmath139 and @xmath140 are trivial in this case . from claim ( ii )
we can know that @xmath109 , @xmath141 and @xmath142 are the only edges of @xmath2 which may penetrate @xmath143 .
but @xmath109 and @xmath142 are excluded from the candidates , because @xmath144 is empty .
recall @xmath123 .
considering the relative positions of the five vertices we know that @xmath141 can not penetrate @xmath143 .
again , combining @xmath145 and @xmath123 with claim ( i ) , we know that the triple @xmath140 is trivial
. therefore we have one more set of trivial triples , that is , either @xmath146 or @xmath147 .
@xmath148 now we observe the relative position of @xmath149 with respect to @xmath105 and @xmath101 .
a key information for this observation is @xmath150 and @xmath90 , that is , @xmath105 dose not penetrate @xmath110 and the vertex @xmath74 belongs to @xmath151 .
firstly note that @xmath152 should not be contained in @xmath153 .
otherwise , since @xmath154 , our key information implies that the vertices @xmath125 and @xmath74 belong to the same half space with respect to @xmath84 .
but it is contradictory to claim ( ii ) .
therefore @xmath152 should belong to the shaded region in figure [ fig4]-(c ) , which implies that @xmath155 intersect both @xmath116 and @xmath156 .
in fact @xmath105 should pass over @xmath149 , because the vertex @xmath74 belongs to @xmath151 , that is , @xmath74 is above @xmath110 with respect the @xmath97-coordinate , and @xmath105 does not penetrate the triangle . and
@xmath101 should pass under @xmath149 , because the edge passes under @xmath111 but does not pierce @xmath110 .
in addition it can be seen that @xmath111 passes under @xmath157 , because the edge penetrates @xmath158 and passes over @xmath101 .
figure [ fig4]-(d ) shows the resulting image @xmath159 .
the observation in the above helps us find another trivial triple @xmath160 . from claim ( ii )
it can be known that @xmath128 , @xmath109 and @xmath105 are the only edges which may penetrate @xmath161 . but looking into figure [ fig4]-(d ) we can exclude the first two candidates . and also the third is excluded , because @xmath105 passes over both @xmath149 and @xmath102 .
now we exclude all hamiltonian cycles of @xmath2 which contain any triple in @xmath162 as consecutive three vertices , because such cycles are not figure-8 as mentioned in the previous section . in the case
that @xmath164 , there remain three heptagons : @xmath165 figure [ fig5 ] shows the images of @xmath166 and @xmath167 under @xmath99 which are constructed by using @xmath119 . as seen in the figure @xmath166 is a trefoil . to see that also @xmath167 is a trefoil , perturb the heptagon slightly .
then , since the vertex @xmath74 is above the vertex @xmath73 , we obtain another image representing a trefoil knot . in the case that @xmath168 , seven heptagons remain : @xmath169 figure [ fig6 ] shows their images under @xmath99 .
the polygons @xmath170 , @xmath171 and @xmath172 were perturbed slightly so that the vertices @xmath73 and @xmath74 are projected to different points . as shown in the figure
@xmath173 is a trefoil knot , and @xmath170 can be isotoped to a polygon representing trefoil . for @xmath171
we need to recall @xmath92 and @xmath100 , which implies that @xmath174 passes over both @xmath101 and @xmath102 .
hence @xmath101 and @xmath102 can be isotoped so that the resulting heptagon has exactly two double points with respect to @xmath99 .
therefore @xmath171 is a trivial knot .
also @xmath175 can be isotoped to a octagon which has three double points , which implies that it is either a trefoil knot or a trivial knot .
similarly @xmath172 is not a figure-8 knot . in this section
we suppose that @xmath36 is of type - ii . and @xmath176 is assumed to be @xmath177 .
note that the table type - ii is almost identical with type - i .
the only difference is that @xmath178 is @xmath106 in type - ii .
therefore , following the procedure in the previous section , we can find out a set of trivial triples @xmath179 which is almost same with the set @xmath162 .
the only difference is that @xmath179 does not include @xmath180 .
@xmath181 also in the same way with the previous section , the image @xmath182 can be obtained .
we observe the image more concretely .
if @xmath152 does not belong to @xmath153 , then @xmath155 should intersect @xmath116 and @xmath156 as shown in the left of figure [ fig7]-(a ) .
since @xmath101 does not penetrate @xmath110 and passes under @xmath111 , the edge should pass under @xmath149 .
in addition @xmath105 should pass under @xmath149 so that @xmath178 is @xmath106 , because @xmath183 is @xmath50 , that is , the vertex @xmath74 is above the plane @xmath84 with respect to the @xmath97-coordinate .
the right of figure [ fig7]-(a ) shows the resulting image of @xmath184 .
figure [ fig7]-(b ) illustrates the case that @xmath152 belongs to @xmath153 . in the case
that the vertex @xmath40 belongs to @xmath135 , after throwing away hamiltonian cycles which have any triple in @xmath179 as consecutive three vertices , we have three remaining heptagons @xmath33 , @xmath185 and @xmath186 .
construct their images by using @xmath119 .
then , irrespective of the position of @xmath152 , the resulting images are same with those in the previous section .
therefore the last two polygons can be shown to be trefoil knots . in the case
that @xmath136 there remain nine heptagons : @xmath187 figure [ fig8 ] shows their images under @xmath99 which are constructed by using figure [ fig7]-(a ) after slight perturbation if necessary .
it is easy to see that @xmath188 , @xmath189 and @xmath190 are not figure-8 . for @xmath191 , @xmath192 and @xmath193
we need to recall that @xmath178 is not zero .
this implies that the edge @xmath149 does not penetrate @xmath194 .
therefore , since the edge passes over @xmath105 , it should pass over @xmath121 . from the resulting images we know that the three heptagons can be isotoped to other polygons which have exactly three double points .
the images obtained from [ fig7]-(b ) represent the same knot types with those from figure [ fig7]-(a ) . in this section
@xmath36 is of type - iii and @xmath176 is assumed to be @xmath177 .
using the table type - iii we attempt to construct the image of @xmath36 under the projection @xmath99 . since @xmath92
, the edges from @xmath106 to @xmath29 should be projected as illustrated in figure [ fig9]-(a ) .
note that @xmath100 and @xmath195 , that is , the vertex @xmath107 belongs to @xmath196 .
hence the vertex @xmath107 should be projected into either the region @xmath197 or @xmath198 .
figure [ fig9]-(b ) shows the image @xmath119 when the vertex @xmath107 is projected into the region @xmath197 .
since @xmath183 is @xmath50 , the vertex @xmath40 should be projected into the shaded region . and since @xmath199 is @xmath106 , @xmath111 should intersect both @xmath101 and @xmath102 . in fact the edge should pass over @xmath101 , because @xmath200 is @xmath106 and @xmath101 already passes over @xmath103 . combing this with @xmath77 we know that @xmath111 passes under @xmath102 .
in addition the edge should pass under @xmath157 , because @xmath201 is @xmath106 .
when the vertex @xmath107 projects into the region @xmath198 , in a similar way , we can construct the image @xmath119 as illustrated in figure [ fig9]-(c ) .
one thing to be noted in this case is that @xmath129 should pass under @xmath101 because the edge passes under @xmath105 but @xmath202 is @xmath203 .
isotoping the resulting image it can be seen that @xmath36 is a trivial knot , a contradiction .
therefore this case can not happen . from the image
@xmath119 we can find a set of trivial triples : @xmath204 see figure [ fig9]-(b ) .
it is obvious that @xmath138 , @xmath205 and @xmath206 are trivial .
also the figure shows that @xmath87 separates @xmath88 from @xmath89 , and @xmath207 separates @xmath208 from @xmath209 , which implies that the next twelve triples are trivial .
the edge @xmath210 is the only edge which may penetrate @xmath211 .
for @xmath212 , @xmath105 is the only possible edge .
but these possibilities are excluded , because @xmath213 is @xmath50 . to find more trivial triples we observe another projected image of @xmath36 .
let @xmath214 be a plane which is orthogonal to @xmath215 .
and @xmath216 is the projection map such that the vertex @xmath40 is above the vertex @xmath106 with respect to the @xmath97-coordinate .
then @xmath111 is projected to a point in @xmath217 as illustrated in figure [ fig10]-(a ) . since @xmath183 is @xmath50 , the vertex @xmath74(_resp .
_ @xmath73 ) is projected into the upper(_resp .
_ lower ) half plane with respect to the line passing through @xmath218 .
hence @xmath219 and @xmath123 imply that @xmath74 belongs to the region @xmath220 , and @xmath73 to @xmath221 respectively . from the figure four more trivial triples
are found : @xmath222 after excluding all hamiltonian cycles which have any of trivial triples in @xmath223 as consecutive three vertices , there remain only three heptagons : @xmath224 figure [ fig10]-(b ) and ( c ) show the images of @xmath225 and @xmath226 under @xmath99 respectively .
it is clear that their knot types are not figure-8 . | in 1983 conway and gordon proved that any embedding of the complete graph @xmath0 into @xmath1 contains at least one nontrivial knot as its hamiltonian cycle .
after their work knots ( also links ) are considered as intrinsic properties of abstract graphs , and numerous subsequent works have been continued until recently . in this paper
we are interested in knotted hamiltonian cycles in linear embedding of @xmath0 .
concretely it is shown that any linear embedding of @xmath0 contains at most three figure-8 knots as its hamiltonian cycles . |
the evolutionary relations among subclasses of magnetic and non - magnetic cataclysmic variables ( mcvs and cvs respectively ) have been studied for quite some time .
nevertheless , our understanding of these relations remains incomplete , caused largely by serious selection effects which prevent completeness of the samples .
here we consider the future evolution of the peculiar mcv ae aqr .
its future spin evolution and equilibrium states could clear much of the current confusion , as this binary has been identified as member of a potentially large group of post - thermal timescale mass transfer systems ( schenker et al . , 2002 ) .
one consequence of this is the prediction of a relatively large population of mcvs at long orbital periods .
such a population has not been detected .
we consider the possibility that this could be because these long period mcvs occupy spin equilibria which significantly hamper accretion onto the white dwarf ( wd ) , causing them to have been overlooked or misclassified .
we first consider the origin of cvs ( fig . 1 ) .
the progenitor of the wd starts off as the more massive star in a relatively wide binary , which has evolved through various stages of ( single ) stellar evolution forming a degenerate core in its centre . during the agb phase
the star swells and fills its roche lobe .
a subsequent phase of common envelope evolution leaves the system with the exposed core ( pre - wd ) and the largely unaffected secondary in a close orbit . in order to be a future
cv this system has to become semi - detached .
this can be achieved by angular momentum loss ( aml ) shrinking the orbit or nuclear evolution increasing the secondary s radius .
the latter channel can lead to a large fraction of systems with a secondary more massive than the wd primary leading to thermal - timescale mass transfer ( ttmt ) .
examples of ttmt evolution are the various model tracks for ae aqr presented in schenker et al.(2002 ) .
in contrast to standard cvs ( which have low mass donors and stable , aml - driven mass transfer ) this subclass passes through an initial phase of high mass transfer prior to become an aml - driven cv . if we consider cvs which have suffered ttmt , we expect there to be differences to standard cvs at the onset of mass transfer : ( i ) the wds are likely to have large rotation rates and masses .
this is a consequence of the high mass transfer rates during the supersoft phase ( during ttmt ) which allows significant accretion onto the wd .
( ii ) instead of low - mass ms stars , the donors are the evolved cores of more massive stars .
their different internal structure , manifesting itself e.g. in a different mass - radius exponent @xmath0 , leads to smaller @xmath1 at the same @xmath2 , and even to different mass transfer rates .
( iii ) the magnetic fields on either or both the donor and the wd may be different ( cf . cumming , 2002 , and this volume , on @xmath3field suppression in rapidly accreting wds ) . the majority of known mcvs ( fig . 2 )
are either polars ( mostly below the period gap ) or intermediate polars ( ips , above the gap up to @xmath4 h ) .
there are only a few , mostly peculiar systems longwards of @xmath5 h , one of which is ae aqr . if many other systems like it have passed through a similar phase , where do they appear on the period distribution ? this lack of magnetic systems between 6 and 10 h we termed the `` ip gap '' , well aware of the sparse observational data to establish its existence ( cf .
recently reheated discussions on evidence for a period gap in mcvs ) .
we use a simplified description for the magnetic drag on gas streaming past a spinning wd ( king , 1993 ) , where the magnetic acceleration @xmath6 is proportional to the velocity shear @xmath7 . after splitting off the radial power dependency in the drag and replacing the @xmath3field and flow velocities with the wd rotation and keplerian motion
respectively we obtain @xmath8 different interaction models are reduced to variations in the drag coefficient @xmath9 and the radial power @xmath10 . propelling and accreting states are possible depending on the wd spin , which determines the sign of the velocity shear . extending the usual definition of the magnetospheric radius ( e.g. warner , 1995 ; frank et al . , 2002 )
we compare the timescale of the magnetic interaction to the dynamical timescale ( in the disk - less case , i ) and to the viscous timescale ( in the truncated disk case , ii ) .
3 shows the resulting critical radii ( shaded for @xmath11 , and an additional set of lines for @xmath12 ) .
for @xmath11 there is a critical spin period @xmath13 marking the maximum of the propelling branch in case i. such a maximum exists for all @xmath14 for both cases .
additionally the asymptotic radius of the accreting branch in the limit of slow rotation can be written as @xmath15 which corresponds to the radius where the stream becomes threaded in an ordinary polar .
both of these expressions describe case i where the mass transfer stream is specified by @xmath16 ( in @xmath17 ) , a stream width @xmath18 ( in @xmath19 ) , and the sound speed at the l1 point @xmath20 ( in @xmath21 ) .
the wd mass is @xmath22 ( in @xmath23 ) and its magnetic moment is @xmath24 ( in @xmath25 ) .
the corresponding expressions for case ii are @xmath26 and @xmath27 with @xmath28 describing the usual disk viscosity .
spin equilibria ips are achieved when torques on the wd cancel .
comparing @xmath29 to various other important radii identifies such states ( warner , 1995 ; king & wynn , 1999 ) .
we are now able to analyse the various spin equilibria in mcvs provided we have good measurements of the periods ( easy ) , magnetic moments ( difficult ) , masses and mass transfer rates ( even more so ) .
results of our most recent simulation are presented in fig .
4 . extending previous attempts ( wynn et al . , 1997 ) to model the current status of ae aqr
, this calculation follows the spin evolution through the next few @xmath30 . during this time
the spin - down continues at roughly the current rate .
we can clearly see that at @xmath31 a truncated disk forms with a large inner hole of @xmath32 .
we compare this result in fig . 5 to an analysis using critical radii .
in the left - hand panel the situation for ae aqr in its current state is shown : at the observed spin period of 33 sec and a strong propeller ( lower curve , as labelled ) , @xmath29 is large enough to prevent disk formation in the circularization region .
as the wd slows down , a disk will form and the effective @xmath29 will switch to the truncated disk case ( pair of curves for hot and cold @xmath28 ) . as long as this happens at @xmath33 ( given by eq .
( 4 ) ) , the next stage of spin evolution is a weak propeller in a truncated disk . in the final state of the computation in fig .
4 we indeed find @xmath34 . for @xmath35 a stable equilibrium is possible where the propeller slows the wd down during quiescence , while it is spun up by an equal amount during outbursts .
this equilibrium can be maintained over secular timescales .
such a system will not be visible as a mcv during its rather long quiescent stage between outbursts .
we can now construct a sequence of diagrams for ae aqr with parameters taken from the post - ttmt evolution by schenker et al .
the right - hand panel of fig .
5 ( showing only the truncated disk pair of curves ) is taken at @xmath36 : the @xmath29 curves are shifted upwards along the corotation radius , i.e. the central hole has grown to the point that spin equilibrium is impossible . further slow down of the wd ends the weak propeller phase , and _ for the first time _
ae aqr will appear as a normal ip : with a typical @xmath37 ratio and @xmath38 in the range where the bulk of ips are found ( fig . 2 ) .
we also identified a potentially stable configuration for a strong ( disk - less ) propeller .
the velocity of a stream reaches its maximum value near the point of closest approach to the wd , the point at which the rotational velocity of the magnetic field is lowest .
so for given @xmath38 and @xmath39 a critical spin period @xmath40 can be found at which the sign of the drag force between the stream and the rotating wd field changes . as the force depends strongly on radius even a small inversion may be sufficient to balance the angular momentum transferred from the wd to the stream while further away .
this may lead to a stable propeller , i.e. a situation where the mass transfer stream is leaving the system yet _ the wd spin does not change_. in order to test this new idea , we have performed sph simulations to investigate the possibility of a stable strong propeller . as can be seen in fig . 6
, it is possible to construct a strong propeller at this critical spin state .
near the point of closest approach the velocity shear reverses and the stream returns angular momentum to the wd
. however , it would appear that this can only happen at spin periods well above @xmath41 found for ae aqr .
in summary both the critical radius analysis and numerical work indicate that ae aqr will continue to be a strong propeller for a short time .
afterwards it is likely to become a weak propeller system with a truncated disk .
at least theoretically , a stable strong propeller is possible , although most probably not feasible for ae aqr . as long as the system can avoid accreting in a stable , ip - like manner , any descendant of ae aqr may easily be overlooked as a mcv .
overall we conclude that ttmt evolution forces us to re - evaluate ideas about magnetic binary evolution .
the various spin states of magnetic wds can lead to drastically different behaviour of otherwise similar systems .
in particular , stable states may exist with or without a disk , propelling or accreting .
not all of these states would currently be considered to be magnetic systems on observational grounds . | we investigate the spin evolution of the unusual magnetic cv ae aqr . as a prototype for a potentially large population of cvs subject to a thermally unstable phase of mass transfer , understanding its future is crucial .
we present a new definition of the magnetospheric radius in terms of the white dwarf s spin period , and use this along with numerical simulations to follow the spin evolution of ae aqr .
we also present preliminary sph results suggesting the existence of a stable propeller state .
these results highlight the complexity of mcvs and may provide am improved understanding of the evolution of all types of cvs .
# 1_#1 _ # 1_#1 _ = # 1 1.25 in .125 in .25 in |
the extreme ultraviolet ( euv ; 150 1200 ) waveband is replete with emission lines formed at temperatures that range from several times @xmath1 k to several times @xmath2 k , and is thus well suited for studying numerous features and phenomena that occur in the solar atmosphere .
like most solar euv instruments , the solar euv research telescope and spectrograph ( serts ) sounding rocket payload observed only part of the euv waveband .
serts was designed to improve on earlier solar euv instrumentation by ( 1 ) retaining the stigmatic imaging capability of the nrl spectroheliograph on _ skylab _ @xcite but with a spatial selection capability that reduced image overlap , and ( 2 ) achieving high spectral resolution .
the version of serts flown in 1989 ( serts-89 ; * ? ? ?
* ) gathered euv radiation with a grazing incidence telescope ( as did subsequent serts flights ) , and carried a standard gold - coated toroidal diffraction grating .
serts-89 observed hundreds of first - order emission lines between 235 and 450 , as well as dozens of second - order lines between 170 and 225 @xcite .
the version of serts flown in 1991 ( serts-91 ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) and 1993 ( serts-93 ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) incorporated a multilayer - coated diffraction grating that enhanced the instrument s first - order responsivity over that of the gold - coated grating by factors up to nine .
the version flown in 1995 ( serts-95 ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) incorporated a multilayer - coated grating that enhanced the instrument s second - order ( @xmath3 170 225 ) responsivity , thus bringing out many lines that had not been seen during any of the previous flights , and allowing them to be measured with the highest spectral resolution ( @xmath3 30 m ) ever achieved for spatially resolved active region and quiet - sun spectra in this wavelength range . for the serts flights of 19891995 data
were recorded on euv - sensitive photographic film that was developed and digitized after each flight . for the flights of 19891993
the relative radiometric calibration was derived by combining laboratory measurements of component responsivities with insensitive ratios " ( solar emission line pairs whose intensity ratios are insensitive to variations in the electron density and temperature ) , while for the flight of 1995 the relative radiometric calibration was derived strictly by the insensitive ratio method . in all four cases the relative radiometric calibration was placed onto an absolute scale by requiring the sum of the he 303.78 and si 303.32 quiet - sun intensities to match the values reported by @xcite based on earlier observations from satellites and sounding rockets . in @xcite the summed intensity is 7115 ergs @xmath4s@xmath5sr@xmath5 at disk center and brightens by about 30% toward the limb .
the absolute radiometric calibration of serts derived this way was estimated to be accurate within a factor better than 2 .
the version of serts flown in 1997 ( serts-97 ; * ? ? ?
* ; * ? ? ?
* ) incorporated the same multilayer - coated toroidal diffraction grating that was flown in 1991 and 1993 , but recorded spectrographic data on an intensified ccd - detector .
its spectral bandpass was 299 353 , with an instrumental resolution ( fwhm ) of 115 m .
an end - to - end radiometric calibration of serts-97 was performed after the flight at rutherford - appleton laboratory ( ral ) , uk , in the same facility used to calibrate the coronal diagnostic spectrometer ( cds ; * ? ? ?
* ) on the _ solar and heliospheric observatory _
( _ soho _ ) spacecraft and the extreme - ultraviolet imaging spectrometer ( eis ; * ? ? ?
* ) aboard the _ hinode _ satellite @xcite , and using the same euv light source re - calibrated by physikalisch - technische bundesanstalt ( ptb ) against the synchrotron radiation of bessy i ( a primary radiation standard ) .
this was the first time that the serts instrument underwent an absolute radiometric calibration while fully assembled .
the uncertainty on the serts-97 absolute radiometric calibration was 15% .
coordinated , cospatial , time - invariant serts-97 and cds spectra were used to carry out an intensity cross - calibration that yielded an improved responsivity curve for the cds normal incidence spectrometer s nis 1 ( 308 380 ) waveband @xcite .
the importance of a reliable radiometric calibration for serts , cds , eis , or any euv instrument can not be over emphasized . without it , quantitative analyses of properties like temperature , density , emission measure , and element abundances , along with the physical interpretations that emerge from studies of those properties , are impossible .
the extreme ultraviolet normal incidence spectrograph ( eunis ) is the successor to serts .
eunis contains two independent but co - pointing spectrographs ( each with a normal incidence telescope ) , one of which covers a short wavelength ( sw ; 170205 ) channel while the other covers a long wavelength ( lw ; 300370 ) channel .
the instrument and its end - to - end absolute radiometric calibration is described below .
spectra obtained from the 2006 flight ( eunis-06 ) were used to investigate a bright point @xcite , a cool transient brightening @xcite , and transition region velocity oscillations @xcite , as well as to derive a calibration update for cds nis 1 @xcite .
eunis was last flown on 2007 november 6 , when the solar disk contained no active regions .
eunis-07 obtained coordinated observations of quiet - sun areas near disk center with both cds and eis . in what follows
we derive updates to the cds and eis absolute radiometric calibrations based on these coordinated observations .
section 2 describes the observations and data reduction ;
3 discusses the absolute radiometric calibration of the eunis-06 and eunis-07 long wavelength channel ; in 4 we derive the calibration update for cds nis with eunis-07 ; 5 presents the absolute radiometric calibration of the eunis-07 sw channel ; in 6 we derive the calibration update for both eis channels with eunis-07 ; and in 7 we discuss and summarize our conclusions .
eunis-07 was launched from white sands missile range , new mexico , at 18:00 ut on 2007 november 6 .
it has a payload mass of 473 kg and reached an apogee of 303 km , spending 369 s above 150 km where low euv extinction permits science - quality data .
a total of 257 exposures , each containing solar spectra and images , were recorded between 1802 and 1808 ut with a frame rate of 1.3 s. eunis observed the quiet region with bright points near the solar disk center .
coordinated observations were obtained with the _
hinode_/eis , the euv imaging telescope ( eit ; * ? ? ?
* ) and the cds aboard _ soho _ , as well as the _ transition region and coronal explorer _ ( @xmath6 ; * ? ? ?
we coalign the data from different instruments using a cross - correlation method by taking eit images as reference .
eunis comprises two independent , spatially co - aligned telescope / spectrographs of identical optical design , one covering euv lines between 300 and 370 seen in first order ( lw channel ) , and a second covering lines between 170 and 205 seen in second order ( sw channel ) .
each telescope is a 110-mm diameter off - axis parabola that forms a real image on a precision slit formed from a single - crystal silicon wafer using a technique developed at gsfc .
the slit image is dispersed , magnified and reimaged by a toroidal grating onto the entrance face of a microchannel plate ( mcp ) intensifier .
each slit has a dumbbell " configuration .
the center 660@xmath7 is a conventional narrow slit ( 2@xmath7 ) . above and below the narrow
slit is a 150@xmath8200@xmath7 lobe " that acts as a slitless spectrograph to produce monochromatic context images in strong emission lines .
these images have proven to be extremely useful for coaligning eunis spectra with other instruments .
eunis has a designed first - order spectral dispersion of 25 m pixel@xmath5 and a designed spatial scale of [email protected] pixel@xmath5 @xcite . for the flight of eunis-06
, its optics limited the actual spatial resolution to about 5@xmath7 and the measured spectral resolution was @xmath3200 and @xmath3100 m fwhm in the lw and sw channels . for the flight of eunis-07 , the lobe - slit fields of view ( fovs ) were oriented north - south near the center of the solar disk . in this flight
all observations were made in scanning mode .
the solar image was scanned perpendicular to the slit at a continuous rate of about 2.4@xmath7 s@xmath5 with full spectral images on all detectors recorded every 1.3 s ( maximum frame rate ) .
pointing began at nominal field center , scanned eastward to @xmath050@xmath7 , reversed , scanned westward to @xmath950@xmath7 , and returned to center .
this operation was repeated until the door was closed , giving almost 7 complete 660@xmath8100@xmath7 spectroheliograms in addition to 253 lobe images in lw channel and 256 in sw channel .
all of the raw data were processed with several routine adjustments , including dark image subtraction , flat - fielding and non - linearity correction before being converted from the recorded data numbers ( dn ) into relative exposure units ( reu ) , which are then used in all subsequent analyses .
eis has both imaging ( 40@xmath7 and 266@xmath7 slots ) and spectroscopic ( 1@xmath7 and 2@xmath7 slits ) capabilities , in the short wavelength ( sw ) range of 170@xmath0210 and the long wavelength ( lw ) range of 250@xmath0290 .
eis has a spectral resolution of about 55 m and a spatial resolution of about 34@xmath7 per pixel .
its spectroscopic mode can operate in a rastering mode ( repeated exposures while scanning over the observation target ) or a sit - and - stare mode ( repeated exposures at the same spatial location ) .
the _ hinode _ spacecraft tracks the solar rotation .
the eis observations were conducted on 2007 november 6 using eis study 209 _ eunis_eis_cross_calibration_. two rastered images were obtained using 2@xmath7@xmath10512@xmath7 slit at 60 or 61 positions with 50 s exposures .
one began at 17:09:41 and ended at 18:00:43 ut prior to the eunis flight , and the other from 18:02:41 to 18:54:34 ut started during the eunis flight .
the eis sequence includes 5 spectral windows in sw covering nearly the full spectrum and 17 windows in lw covering selected spectral lines .
the raw data were processed by the standard routine _
eis_prep _ provided by solarsoftware ( ssw ) to remove detector bias and dark current , hot pixels , and cosmic rays , and to make absolute radiometric calibration .
the eis slit tilt and orbital variation in the line centroids were also removed from the data .
the pointing between eis sw and lw detectors has offsets of 2@xmath7 in the x - direction and about 18@xmath7 in the y - direction @xcite .
llccccccc 303.78 & he & 4759 & 5244 & 4484 & 4290 & 0.91 @xmath11 0.13 & 1.06 @xmath11 0.15 & 1.11 @xmath11 0.16 + 313.76 & mg & 25.2 & 31.7 & 25.9 & 14.1 & 0.79 @xmath11 0.11 & 0.97 @xmath11 0.14 & 1.79 @xmath11 0.25 + 314.31 & si & 25.1 & 36.1 & 30.0 & 24.1 & 0.70 @xmath11 0.10 & 0.84 @xmath11 0.12 & 1.04 @xmath11 0.15 + 315.01 & mg & 62.3 & 37.6 & 31.7 & 22.0 & 1.66 @xmath11 0.23 & 1.97 @xmath11 0.28 & 2.83 @xmath11 0.40 + 316.20 & si & 39.2 & 44.7 & 38.7 & 26.6 & 0.88 @xmath11 0.12 & 1.01 @xmath11 0.14 & 1.47 @xmath11 0.21 + 319.81 & si & 60.5 & 48.6 & 45.5 & 26.4 & 1.24 @xmath11 0.18 & 1.33 @xmath11 0.19 & 2.29 @xmath11 0.32 + 341.91 & si & 15.0 & 17.0 & 15.4 & 13.9 & 0.88 @xmath11 0.12 & 0.97 @xmath11 0.14 & 1.08 @xmath11 0.15 + 345.04 & si & 34.0 & 66.8 & 60.9 & 46.5 & 0.51 @xmath11 0.07 & 0.56 @xmath11 0.08 & 0.73 @xmath11 0.10 + 345.67 & fe & 20.6 & 14.7 & 13.4 & 15.7 & 1.40 @xmath11 0.20 & 1.54 @xmath11 0.22 & 1.31 @xmath11 0.19 + 347.34 & si & 24.0 & 27.6 & 25.6 & 22.2 & 0.87 @xmath11 0.12 & 0.94 @xmath11 0.13 & 1.08 @xmath11 0.15 + 352.58 & fe & 30.3 & 19.9 & 20.0 & 20.4 & 1.52 @xmath11 0.22 & 1.51 @xmath11 0.21 & 1.49 @xmath11 0.21 + 368.11 & mg & 285.8 & 262.4 & 252.4 & 172.1 & 1.09 @xmath11 0.15 & 1.13 @xmath11 0.16 & 1.66 @xmath11 0.23 the cds includes a normal incidence spectrometer ( nis ) that can be used to obtain stigmatic euv spectra within its 308@xmath0381 ( nis 1 ) and 513@xmath0633 ( nis 2 ) wavebands along its 240@xmath7 long slit .
several slit widths are available , the most commonly used being 4@xmath7 . the instrument can be operated in a sit - and - stare mode or a rastering mode . in the latter case ,
the cds scans a region of the sun from the west to east without compensation for solar rotation , thus the actual fov of a rastered spectroheliogram is stretched out in the x - direction when the targeted region is located on the solar disk . in this study , the cds rastered image was observed from 17:00:28 to 19:01:00 ut , consisting of 60 pointing positions .
the fov is 260@xmath12 .
the pointing difference between cds nis 1 and nis 2 has been corrected when applying _
mk_cds_map _ in ssw idl library .
we first built up eunis rastered slit - lobe images in selected spectral lines for lw and sw channels using all exposure frames .
figure [ fgfov ] shows the lw he 304 slit - lobe image and the sw composite image from three coronal lines , fe 174.5 , fe 177.2 and fe 180.4 .
we then determined the pointing , roll angle , and actual spatial scale of eunis slit images by coaligning them with eit images in 304 and 195 passbands using the method as described in @xcite . for the eunis lw images we obtained the actual pixel size of [email protected] pixel@xmath5 and the roll angle of [email protected] counterclockwise from the north .
for the sw images we obtained the actual pixel size of [email protected] pixel@xmath5 and the roll angle of [email protected] .
the measurements show that the lw and sw channels have nearly the same pixel size and roll angle , and they were well co - pointed in the direction perpendicular to the slit with a small offset of [email protected] in solar x - direction and an offset of [email protected] in solar y - direction .
figure [ fgfov ] shows the coaligned eunis-07 sw and lw slit - lobe images , indicating that their fovs are mostly overlaid in the slit part .
figure [ fgefv ] illustrates their cospatial rastered slit images in si 347 line of lw channel and fe 188 line of sw channel .
the pointing of eis lw band was determined by coaligning the he 256 rastered image obtained during 18:0218:54 ut with the average image of eit 304 at 18:01 and 18:09 ut . the pointing of eis sw band was determined by coaligning the fe 195 raster image with the average image of eit 195 at 17:55 and 18:22 ut , where the average was made after the two eit images were rotated to a common time .
the measured pointings for both bands are the same considering the existing offsets between them .
figure [ fgcfv ] illustrates the accurate coalignments between eunis-07 and eis in both bands .
the pointing of cds nis was determined by coaligning the he 304 ( second order ) rastered image observed during 17:00@xmath019:00 ut with the eit 304 image at 18:01 ut .
figure [ fgcdsmap ] illustrates a good coalignment between eunis and cds nis as seen in he 304 and mg 368 rastered images .
comparisons between the fovs of coaligned eunis , eis , and cds nis are shown in figure [ fgfov ] .
radiometric calibration of the eunis-06 lw channel was carried out in august 2006 and that of the eunis-07 lw channel was in may 2008 .
they were performed at the ral in the same facility and using the same euv light source as was used for preflight calibrations of cds @xcite .
recalibration of the german ptb light source against the primary euv radiation standard of bessy - ii ( _ english : _ berlin electron storage ring society for synchrotron radiation ) in march 2007 showed that it had remained stable within its 7% uncertainty .
the end - to - end calibration of eunis-07 was made using the he 304 line at 88 individual locations , and using 11 distinct ne features between 300 and 370 at 176 individual locations covering the instrument s entrance aperture .
the aperture - averaged response at each wavelength was combined with the known source flux and with geometric factors , such as aperture area and pixel size , resulting in the absolute eunis responsivity within a total uncertainty of 10% over its full lw bandpass .
the spectral responsivity measures the recorded data numbers per photon flux in ergs @xmath4s@xmath5sr@xmath5 .
the measured absolute responsivities of eunis-07 and eunis-06 as a function of wavelength are shown in figure [ fglwc ] .
the shape of the response curve is mainly due to the multilayer coating on the spectrograph grating which is applied to enhance its euv efficiency .
error bars reflect the total range of values measured from the various entrance aperture positions .
the fiber - optic coupling between the mcp and three active pixel sensors ( aps ) arrays in each optical channel leads to differences in the overall responsivity of each aps array relative to the other two . after correcting for these factors ( @xmath14 ) which can be measured in laboratory , the response curve ( @xmath15 ) in the range of 300370 of three aps arrays
was obtained by a least - square parabolic fit to the measurements using @xmath16 where @xmath17 is the fitting function with @xmath18=335 and @xmath15 is in units of reu ( erg @xmath4sr@xmath5pixel@xmath5)@xmath5 , where the pixel unit is referred to as ` spectral pixel ' which is equal to 25 m for first order spectrum . for eunis-07 we obtained the parameters , @xmath19=([email protected])@xmath1010@xmath20 , @xmath21=([email protected])@xmath1010@xmath20 , @xmath22=@xmath0([email protected])@xmath1010@xmath23 , while for eunis-06 obtained @xmath19=([email protected])@xmath1010@xmath5 , @xmath21=([email protected])@xmath1010@xmath20 , @xmath22=@xmath0([email protected])@xmath1010@xmath23 .
the shape of the response curve for eunis-07 has little change compared to that of eunis-06 ( figures [ fglwc](a ) and [ fglwc](b ) ) , but its absolute responsivity has dropped by a factor of [email protected] ( or decreased by @xmath331% ) ( figure [ fglwc](c ) ) .
there was a slight shift in wavelength covered by eunis-07 relative to eunis-06 due to a laboratory adjustment ( see fig .
[ fglwc](b ) ) .
validation of the laboratory calibration of the eunis-06 lw channel relevant to the shape of its response has been verified by checking its relative calibration using the density- and temperature - insensitive line ratio method @xcite .
therefore , no change in the shape of the response curve for eunis-07 supports the validation of its laboratory calibration . in other words , because the insensitive ratio method confirmed the eunis-06 calibration , and the shape of the eunis-07 response curve remained constant relative to eunis-06 , there is no need to repeat the insensitive ratio test .
llccccc fe & 345.74 & 1.00 @xmath11 0.00 & ... & 22.90@xmath24 2.29 & ... + & 174.53 & 21.08 @xmath11 3.04 & 1.21 @xmath11 0.12 & 482.63 @xmath11 84.70 & 2.51 @xmath11 0.51 & 2.51 @xmath11 0.51 + & 177.24 & 11.59 @xmath11 1.54 & 0.81 @xmath11 0.08 & 265.35 @xmath11 44.13 & 3.05 @xmath11 0.59 & 3.05
@xmath11 0.59 + & 184.54 & 4.97 @xmath11 0.19 & 1.57 @xmath11 0.16 & 113.79 @xmath11 12.18 & 13.75 @xmath11 2.01 & 4.23 @xmath11 0.62 + + fe & 352.66 & 1.00 @xmath11 0.00 & ... & 31.33 @xmath11 3.13 & ... + & 180.41 & 11.44 @xmath11 1.24 & 1.22 @xmath11 0.12 & 358.42 @xmath11 52.86 & 3.40 @xmath11 0.61 & 3.40 @xmath11 0.61 + & 188.23 & 7.87 @xmath11 0.24 & 3.29 @xmath11 0.33 & 246.57 @xmath11 25.78 & 13.33 @xmath11 1.93 & 4.10@xmath11 0.59 + + fe & 364.47 & 1.00 @xmath11 0.00 & ... & 21.04 @xmath11 2.10 & ... + & 192.39 & 1.94 @xmath11 0.06 & 0.40 @xmath11 0.04 & 40.83 @xmath11 4.27 & 9.80
@xmath11 1.42 & 3.01 @xmath11 0.44 + & 193.51 & 4.06 @xmath11 0.15 & 0.93 @xmath11 0.09 & 85.44 @xmath11 9.11 & 10.84 @xmath11 1.58 & 3.33 @xmath11 0.49
the cds nis instrument was calibrated end - to - end at ral , and various underflight calibrations were implemented later on , forming the version 4 of the nis ` standard ' calibration ( implemented on may 21 , 2002 ) .
the history of the nis calibration was described in detail by @xcite .
@xcite used the line - ratio technique to obtain an independent set of responsivities which were consistent , to within a relative 30% , with the ` standard ' values ( see the 2002 book on the soho calibration ) .
the mcp in nis detector is known to have a drop in sensitivity over time due to exposure to solar radiation .
this results in a depression at the core of the lines when one of the 2@xmath7 or 4@xmath7 slits is used ( the so - called _ burn - in _ of the lines ) .
this effect in the nis spectra has been monitored since launch with the nimcp study and is routinely corrected .
the use of the wide 90@xmath7 slit was thought to be the main cause of the responsivity loss over long time - scales @xcite .
an empirical correction was implemented within the standard software , and is still applied by default .
however , @xcite used the same nimcp study to actually show that the burn - in due to the use of the 90@xmath7 slit is a secondary effect , while the responsivity in both nis channels has been steadily dropping .
@xcite provided a wavelength - dependent long - term correction for this drop over the period 1996 - 2010 .
the @xcite long - term corrections , combined with the @xcite responsivities provided good agreement between the irradiances of most cds lines with those measured with the prototype of the solar dynamics observatory ( sdo ) eve instrument , flown on a rocket on april 14 , 2008 .
the eunis lw channel has a wavelength range overlapping with that of cds / nis 1 , so can provide a direct calibration update for it .
the lab - calibrated serts-97 was successfully used to improve the response curve for the cds nis 1 waveband based on their coordinated , cospatial spectra @xcite .
based on observations of active region spectra , eunis-06 provided a new calibration update for cds nis @xcite , which showed an overall decrease in nis 1 responsivity by a factor about 1.7 compared to that of the previously implemented calibration ( version 4 , 2002 ) with the nis standard " long - term corrections , while the responsivity in nis 2 second order at 303@xmath0304 with the nis standard " long - term corrections remained nearly constant ( fig .
[ fgcdscal](a ) ) .
it needs to be pointed out that the measured cds he 304 intensity in @xcite was underestimated by a factor of 2 due to a bug found in the calibration routine , _
nis_calib _ , which was fixed as of 2011 may . in the following
we present the cds underflight calibration update with eunis-07 measurements .
figure [ fgcdsmap ] shows the common fovs between eunis-07 lw and cds nis as illustrated by the rastered images in he 304 and mg 368 , indicating that a quiet region was observed .
figure [ fgcdsspc ] shows their cospatial average spectra , where 202 exposure frames were averaged for eunis-07 .
note that the two regions around the 335 and 360 lines of cds nis 1 spectrum are suffering from depression due to wide - slit burn - in effects from the fe lines , where the long - term correction is more uncertain @xcite .
both the calibrated eunis-07 lw and cds nis spectra ( as well as the calibrated eunis-07 sw and eis spectra shown in figure [ fgsws ] and figure [ fglws ] ) exhibit a significant background particularly in the low - responsivity wavelength range .
this mainly resulted from a combined effect of the instrumental noise and scattered light with the special shape of their response curves . for the eunis spectra , we first divided the full spectral window into several small sections , and then fitted spectral lines by a multiple gaussian function with a linear background using the standard line fitting routine ( _ xcfit.pro _ ) in ssw . in this way
the slowly , smoothly varying ( with wavelength ) `` background '' level only slightly affects the integrated intensities of the fitted line profiles .
measurements of the line intensities from eunis-07 sw and eis sw / lw spectra in sections [ sctswc ] and [ scteis ] were made using the same technique .
the cds line intensities were obtained by fitting multiple broadened gaussian functions to the full spectrum .
the broadened gaussian function is a combination of a gaussian term plus a term describing the wings @xcite .
we chose the broadening with no asymmetry and kept the relative amplitude of the wings equal to a value of 0.8 , but allowed the width of the gaussian term for the line to vary ( within 0.29 and 0.4 in fwhm ) , as well as its centroid and peak values .
all lines were fitted at once , with the background fitted with a polynomial curve .
figure [ fgnisft ] shows the fitted cds line profiles in two selected wavelength ranges .
even with its moderate spectral resolution , cds can resolve the lines that we used for calibration .
the uncertainty caused by the moderate spectral resolution of cds is relatively small compared to that in the background when assuming the common background fit for all spectral lines .
table [ tabcds ] lists 12 strong emission lines observed by both eunis lw and cds ( nis 1 in first order and nis 2 in second order ) . to examine the current cds responsivities and long - term corrections
, we measured the line intensities with three kinds of calibration as used in @xcite . the fourth column ( @xmath25 )
shows the intensities obtained with the current standard cds responsivities and the new long - term corrections derived by @xcite , which has been implemented with the keyword _
alt_slit6 _ in the ssw routine
the fifth column ( @xmath26 ) shows the intensities obtained with the @xcite scaled responsivities ( using _ nis_resp_gdz_2010 _ in ssw ) and the new long - term corrections .
note that differences between the standard cds responsivities and the @xcite scaled ones are small , overall of the order of 10% or so .
the sixth column ( @xmath27 ) shows the intensities obtained with the current standard cds responsivities ( version 4 , 2002 ) and the standard long - term corrections ( using _ vds_calib _ with the keyword _
slit6 _ [ default ] ) . considering long - term corrections
have the relative uncertainty of 20% , combined with the uncertainties in the background and the responsivities ( 20% to 30% see * ? ? ?
* ) , it means a typical uncertainty of 30%40% in measurements of cds line intensities
. ratios of the eunis-07 to various cds line intensities are listed in columns 7@xmath09 and shown in figure [ fgcdscal ] , where errors of the ratios were calculated by assuming 10% uncertainty ( as a lower limit ) in measurements of both eunis and cds line intensities .
the differences are significant between the eunis-07 line intensities and the cds ones applied with standard responsivities and standard long - term corrections , especially for those lines in the wavelength range of 310@xmath0320 .
the nis 1 responsivity ( with standard long - term corrections ) overall decreased by a factor of [email protected] , which is consistent with that derived by eunis-06 . while the eunis-07 line intensities are well consistent with the cds ones applied with new long - term corrections in many lines ( with differences less than 15% ) .
their average ratio is [email protected] in the case of the standard responsivities and it is [email protected] in the case of the @xcite scaled ones .
therefore , the results of eunis-06 and eunis-07 flights confirmed the new long - term corrections for nis 1 response obtained by @xcite , and strongly suggest the application of keyword , _
alt_slit6 _ , in using _
vds_calib _ or _
nis_calib _ to calibrate cds data .
in addition , our measurements show that the differences are small ( on the order of 10% ) between he 304 intensities obtained by eunis-07 in first order and by cds nis-2 in second order applied with the standard or new long - term corrections , indicating that the second - order standard ( version 4 ) responsivity of nis 2 updated by serts-97 @xcite and egs @xcite remains to work well ( with a decrease in uncertainty ) when the standard long - term corrections are applied . finally , we notice that even after the new long - term corrections , the cds intensities for some lines showed large ( @xmath2830% ) difference from the eunis ones
. this could be due to the uncertainties involved in difficulty of background ( scattered light ) estimate for the cds spectrum .
llcccccc fe & 174.54 & 1.281 @xmath11 0.128 & 522.37 @xmath11 52.24 & 0.799 @xmath11 0.080 & 393.03 @xmath11 39.30 & 1.329 @xmath11 0.188 & 1.53e-03 @xmath11 2.16e-04 + fe & 177.24 & 0.799 @xmath11 0.080 & 261.56 @xmath11 26.16 & 1.313 @xmath11 0.131 & 212.94 @xmath11 21.29 & 1.228 @xmath11 0.174 & 5.02e-03 @xmath11 7.10e-04 + fe & 180.39 & 1.261 @xmath11 0.126 & 344.89 @xmath11 34.49 & 5.530 @xmath11 0.553 & 257.12 @xmath11 25.71 & 1.341 @xmath11 0.190 & 1.60e-02 @xmath11 2.27e-03 + fe & 184.54 & 1.617 @xmath11 0.162 & 122.16 @xmath11 12.22 & 8.524 @xmath11 0.852 & 102.33 @xmath11 10.23 & 1.194 @xmath11 0.169 & 6.98e-02 @xmath11 9.87e-03 + fe & 185.22 & 0.484 @xmath11 0.048 & 36.46 @xmath11 3.65 & 3.035 @xmath11 0.303 & 30.48 @xmath11 3.05 & 1.196 @xmath11 0.169 & 8.32e-02 @xmath11 1.18e-02 + fe & 186.60 & 0.335 @xmath11 0.034 & 25.35 @xmath11 2.54 & 3.229 @xmath11 0.323 & 23.41 @xmath11 2.34 & 1.083 @xmath11 0.153 & 1.27e-01 @xmath11 1.80e-02 + fe & 186.88 & 0.318 @xmath11 0.032 & 24.13 @xmath11 2.41 & 3.204 @xmath11 0.320 & 21.93 @xmath11 2.19 & 1.100 @xmath11 0.156 & 1.33e-01 @xmath11 1.88e-02 + fe & 188.23 & 3.515 @xmath11 0.351 & 272.88 @xmath11 27.29 & 39.602 @xmath11 3.960 & 209.95 @xmath11 21.00 & 1.300 @xmath11 0.184 & 1.45e-01 @xmath11 2.05e-02 + fe & 190.04 & 0.676 @xmath11 0.068 & 55.53 @xmath11 5.55 & 12.383 @xmath11 1.238 & 49.60 @xmath11 4.96 & 1.120 @xmath11 0.158 & 2.23e-01 @xmath11 3.15e-02 + fe & 192.40 & 0.444 @xmath11 0.044 & 40.93 @xmath11 4.09 & 10.596 @xmath11 1.060 & 32.49 @xmath11 3.25 & 1.260 @xmath11 0.178 & 2.59e-01 @xmath11 3.66e-02 + fe & 193.51 & 0.959 @xmath11 0.096 & 94.80 @xmath11 9.48 & 26.673 @xmath11 2.667 & 74.79
@xmath11 7.48 & 1.268 @xmath11 0.179 & 2.81e-01 @xmath11 3.98e-02
the insensitive line ratio method was proposed by @xcite as a means of monitoring relative calibration variations of inflight euv spectrometers and was used by @xcite and @xcite to adjust the laboratory calibration curve for serts-89 , serts-91 and serts-93 .
@xcite derived the serts-95 relative radiometric calibration for both the first - order and second - order wave bands with this technique . in the following we used the lab - calibrated eunis-07 lw channel to calibrate its sw channel by means of density- and temperature - insensitive line intensity ratios because the responsivity of its sw channel could not be directly measured at ral .
figure [ fgefv ] shows the common fov between lw and sw channels of eunis-07 as illustrated by the rastered images in coronal lines , si 347.4 from lw and fe 188 ( [email protected] ) from sw .
figure [ fgesp ] shows the spectra averaged over the same spatial area for 202 exposure frames observed from 18:02:14 to 18:06:34 ut .
line intensities of the eunis lw spectrum were measured with the gaussian fit , while those of the sw spectrum with the broadened gaussian fit .
table [ tabswc ] lists three groups of emission lines for fe , fe , and fe .
column ( 3 ) gives theoretical line intensities relative to the one in the wavelength range for lw , which were calculated with the chianti ver .
6 package @xcite .
the line ratios exhibit slight variations with electron density ( @xmath29 ) , so the mean value over a range of @xmath30 was taken , with the uncertainty corresponding to half of the difference between the maximum and minimum values . in each case , the intensity ratios were calculated at the temperature of maximum ion abundance ( i.e. , the lines formation temperature ) , using the ionization equilibrium data ( _ chianti.ioneq_ ) .
column ( 4 ) gives the uncalibrated line intensity of sw spectral lines . in column
( 5 ) the absolute intensity for lw lines was directly measured from the average spectrum , while the absolute intensity of sw lines was derived from the measured lw line intensity and chianti theoretical insensitive line ratio .
column ( 6 ) gives the absolute responsivities , which are the ratio of the measured uncalibrated intensity ( column ( 4 ) ) to the derived calibrated intensity ( column ( 5 ) ) for sw lines .
column ( 7 ) gives the relative responsivities which were derived from the absolute ones by correcting for the relative sensitivity factors ( @xmath14 ) .
figure [ fgucl](a ) shows the relative responsivities and a least - squares parabolic fit on a logarithmic scale . the radiometric calibration response curve ( @xmath15 ) in the range of 170@xmath0205
is shown in figure [ fgucl](b ) , obtained by equations ( [ eqsen ] ) and ( [ eqres ] ) , where @xmath17 is the fitting function with @xmath18=187.5 , @[email protected]@xmath110.04 , @xmath21=@xmath0([email protected])@xmath1010@xmath20 , @xmath22=@xmath0([email protected])@xmath1010@xmath20 , and the relative sensitivity factors for the three aps arrays were determined in laboratory as @xmath31 the @xmath15 has units of reu ( erg @xmath4sr@xmath5@xmath5)@xmath5 .
the shape of the sw response curve of eunis-07 is consistent with that of eunis-06 within the 15% uncertainty , but its responsivity has decreased by a factor of 2 . note that because the responsivity had fallen significantly and the observed region was very quiet , no emission lines were observed by the right asp detector in sw channel which covers the wavelength range from 196 to 207 , including three fe ( 200.0 , 203.7 , 202.0 ) lines that had been used for eunis-06 sw radiometric calibration .
therefore , the responsivities in the wavelength range between 194 and 205 are not directly constrained by observations .
llcccc fe & 345.74 & [email protected] & [email protected] & ... & ... + & 174.53 & 21.07@xmath11 3.04 & 491.98@xmath11 86.37 & 389.96@xmath11 39.00 & 1.26@xmath11 0.25 + & 177.24 & 11.58@xmath11 1.54 & 270.39@xmath11 44.99 & 209.59@xmath11 20.96 & 1.29@xmath11 0.25 + & 184.54 & 4.97@xmath11 0.19 & 116.05@xmath11 12.42 & 101.55@xmath11 10.16 & 1.14@xmath11 0.17 + fe & 352.66 & [email protected] & [email protected] & ... & ... + & 188.23 & 7.87@xmath11 0.24 & 277.42@xmath11 30.85 & 212.92@xmath11 21.29 & 1.30@xmath11 0.19 + & 192.83 & 1.20@xmath11 0.06 & 42.30@xmath11 4.99 & 34.99@xmath11 3.50 & 1.21@xmath11 0.19 + fe & 341.11 & [email protected] & [email protected] & ... & ... + & 188.23 & 28.33@xmath11 4.67 & 303.98@xmath11 58.64 & 212.92@xmath11 21.29 & 1.43@xmath11 0.31 + & 192.83 & 4.31@xmath11 0.77 & 46.25@xmath11 9.47 & 34.99@xmath11 3.50 & 1.32@xmath11 0.30 + fe & 352.11 & [email protected] & [email protected] & ...
& ... + & 192.39 & 3.22@xmath11 0.14 & 42.17@xmath11 11.01 & 32.86@xmath11 3.29 & 1.28@xmath11 0.36 + & 193.51 & 6.75@xmath11 0.34 & 88.39@xmath11 23.18 & 76.74@xmath11 7.67 & 1.15@xmath11 0.32 + & 195.12 & 10.63@xmath11 0.19 & 139.20@xmath11 35.91 & 117.14@xmath11 11.71 & 1.19@xmath11 0.33 + fe & 364.47 & [email protected] & [email protected] & ... & ... + & 192.39 & 1.94@xmath11 0.06 & 40.88@xmath11 5.37 & 32.86@xmath11 3.29 & 1.24@xmath11 0.21 + & 193.51 & 4.06@xmath11 0.15 & 85.54@xmath11 11.37 & 76.74@xmath11 7.67 & 1.11@xmath11 0.19 + & 195.12 & 6.39@xmath11 0.21 & 134.64@xmath11 17.75 & 117.14@xmath11 11.71 & 1.15@xmath11 0.19 + fe & 338.26 & [email protected] & [email protected] & ... & ... + & 186.85 & 9.32@xmath11 1.94 & [email protected] & 24.08@xmath11 2.41 & 1.27@xmath11 0.32 + si & 347.41 & [email protected] & [email protected] & ... & ... + & 261.04 & 0.79@xmath11 0.12 & [email protected] & 19.31@xmath11 1.93 & 1.26@xmath11 0.26 + & 277.28 & 0.48@xmath11 0.07 & 14.83@xmath11 2.6 & 11.63@xmath11 1.16 & 1.28@xmath11 0.26 + & 272.01 & 0.59@xmath11 0.09 & 18.23@xmath11 3.32 & 17.03@xmath11 1.70 & 1.07@xmath11 0.22
we determined the eis radiometric calibration using two methods .
one is based on direct transfer of the radiometric calibration for eunis-07 sw channel , whose wavelength range overlaps that of the eis sw band .
the other is based on the temperature- and density - insensitive line ratio technique , in which one member of each insensitive line pair was observed by eunis-07 lw channel , and the other by eis lw or sw band . in the following
we applied the first method to the eis raster observed after the eunis-07 flight , and applied the second method to the eis rasters observed both before and after the eunis-07 flight .
figure [ fgsws ] shows the cospatial average spectra of eunis-07 sw and eis sw bands over a common fov shown in figures [ fgcfv](c ) and [ fgcfv](d ) .
the eunis sw spectrum was calibrated with the response curve determined in section [ sctswc ] .
the eis spectrum was calibrated using the standard routine , _
eis_prep _ in ssw , which used the pre - launch radiometric calibration based on laboratory measurements @xcite .
the eis average spectrum was fitted using an interactive multi - gaussian ( with a linear background ) fitting routine , _
spec_gauss_eis _ in ssw .
table [ tabeis ] lists the measured calibrated and uncalibrated line intensities for 11 strong emission lines of fe , fe , fe , and fe .
a 10% uncertainty was assumed for the measured eunis and eis line intensities because the statistical errors in the fitting were negligible , while the uncertainties due to absolute radiometric calibration may be up to 20% .
column ( 7 ) gives the eunis - to - eis line intensity ratios ( see figure [ fgecl](a ) ) .
column ( 8) gives the derived eis responsivities which are ratios of the uncalibrated eis to calibrated eunis line intensities ( see figure [ fgecl](b ) ) .
our measurements indicate that the eis responsivity decreased by a factor of [email protected] after a year inflight .
the shape of the derived eis response curve is well consistent with the one measured before its launch .
the eis response curve ( in units of dn ( erg @xmath4sr@xmath5pixel@xmath5)@xmath5 ) derived with a least - squares parabolic fit on a logarithmic scale in the wavelength range of 174@xmath0194 is @xmath32 where @xmath18=185 , @[email protected]@xmath110.03 , @[email protected] , @xmath22=@xmath0([email protected])@xmath1010@xmath20 .
note that the eis responsivity ( r@xmath33 ) derived above is for 2@xmath7 slit , while in the case of 1@xmath7 slit it should be reduced by a factor of 2 . figure [ fglws ] shows the absolutely calibrated spectra for eunis-07 lw and eis lw and sw bands , averaged over a common fov shown in figures [ fgcfv](a ) and [ fgcfv](b ) .
table [ tabrat ] lists seven groups of temperature- and density - insensitive lines from fe , fe , fe , and si .
since the two fe lw lines , 352.66 and 341.11 , and the two fe lw lines , 352.11 and 364.47 , are insensitive line pairs , they were separated forming two groups which include the same sw line members .
column ( 3 ) gives theoretical line intensities normalized to the first member ( eunis lw line ) in each group . in column
( 4 ) the absolute intensity of the eunis lw line was directly measured from the average spectrum , while those of the sw lines were derived using the theoretical ratios . column ( 5 ) gives the absolute intensity of eis sw lines measured from the average spectrum .
column ( 6 ) are ratios of the eunis - derived line intensity to the eis directly measured one , which are shown in figure [ fgrat ] .
our measurements from the eis data set obtained from 18:02 to 18:54 ut indicate that the eis responsivities in both bands decreased by a factor of [email protected] compared to the pre - launch ones measured in laboratory , which are in good agreement with the result obtained using method 1 . in the case using the eis data set observed from 17:09 to 18:00 ut prior to the eunis-07 flight ,
the obtained line ratios ( red symbols shown in figure [ fgrat ] ) indicate that the eis responsivities decreased by a factor of [email protected] , about 5% smaller than that obtained in the former case .
this small difference may be caused by some cooling bright points observed at the southern end of the fov . by averaging the results from two eis data sets
, we derived that the responsivities of two eis bands degraded by a factor of 1.2 ( or by 17% ) after one year flight although the size of the measurement uncertainties is comparable to this decrease .
using coordinated , cospatial spectroscopic observations of a quiet region of the sun , the lab - calibrated eunis-07 lw channel has been directly applied to update the cds nis calibration .
we have found that the measured cds nis 1 line intensities calibrated with the standard ( version 4 ) responsivities with the standard long - term corrections are overall underestimated by a factor of [email protected] due to the improper treatment of the decrease in responsivity , while the eunis-07 measurements are in very good agreement with the new long - term corrections for nis 1 derived by @xcite .
the difference is small ( by about 10% ) between the he 304 intensities measured by eunis-07 in first order and by cds nis 2 in second order with the standard or new long - term corrections .
therefore the eunis-07 and eunis-06 have provided the consistent calibration updates for cds nis 1 in the wavelength range of 301370 and nis 2 in second order at 303304 and confirmed the inflight calibration update by @xcite .
it is worth pointing out both the standard and the new long - term corrections for cds / nis include a wavelength - dependent decrease in responsivity .
a major difference lies in that the standard correction assumed that the brightest lines would be most affected , so applies a correction which is too large for the strongest lines , while too small for the weaker lines .
the new long - term corrections indicate an overall decrease in the nis responsivity at all wavelengths . the lab - calibrated eunis-07 lw channel is also applied to derive the eunis-07 sw channel calibration using a technique based on density- and temperature - insensitive line intensity ratios .
the eis calibration update is performed in two ways .
one is using the direct calibration transfer of the calibrated eunis-07 sw channel .
the other is using the insensitive line pairs , in which one member was observed by eunis-07 lw channel and the other by eis in either lw or sw waveband .
the response curve of eis sw waveband is derived using the first method , which well coincides in shape with the lab - calibrated one prior to launch .
results obtained from the two methods are in good agreement , and confirm ( within the measurement uncertainties ) the eis responsivity measured directly before the instrument s launch .
the measurements also suggest that the responsivity of both eis wavebands decreased by a factor of 1.2 ( or dropped to 83% of their prelaunch value ) after a year flight ( 2007 november 6 ) . the overall uncertainty due to the slightly temperature- and density - sensitivity of theoretical line ratios is less than 10% .
responsivity loss of euv spectrometers is often measured by monitoring the intensities of solar spectral lines from the basal chromospheric - transition region of quiet sun , which are known on average not to vary over the cycle ( e.g. * ? ? ?
a recent eis science nugget by doschek ( 2011 ) reported an e - folding time of 6.4 years for the eis instrument s responsivity by measuring the intensity variation of the he line at 256.32 using the synoptic data between january 2007 and july 2009 . by including quiet sun data near disk center early in the mission before the synoptics started and the synoptic data in 2010 , mariska ( 2011 , private communication ) updated the measurement of e - fold time to be 5.2 years , which gives a decrease of responsivity by a factor of 1.2 for the first year s flight , very well consistent with that derived in the present work .
however , a preliminary investigation of the eis radiometric calibration based on chianti beta v.7 showed that the updated atomic data would reduce the eunis - to - eis line ratios by 10% for fe and si , and thus the degradation of eis responsivity by a factor of about 1.2 derived here may still be within the uncertainties of the eis prelaunch calibration .
in addition , based on the sumer / eis joint campaigns in 2007 march april , @xcite made the sumer
eis relative intensity calibration and found that the two instruments agreed with each other within uncertainties .
lllll eunis-07 & 2007-nov-06 & quiet region & 4960@xmath11496 & this paper + eunis-06 & 2006-apr-12 & quiet region & 4108@xmath11411 & this paper + sdo / eve & 2010 may - dec & full sun & 5670@xmath11190 ( 59@xmath112 ) bl & this paper + prototype - eve rocket & 2008-apr-14 & full sun & 5766 ( 60 ) bl & @xcite + cds nis & 2006 - 2008 & full sun & 5382@xmath11288 ( 56@xmath113 ) & @xcite + cds nis & 1997-may-15 & full sun & 6520@xmath111304 ( 68@xmath1114 ) & @xcite + cds nis & 1998-may-1 & quiet region & 6856@xmath11132 & @xcite + cds nis & 1996 mar - jun & quiet region & 7204@xmath118 & @xcite + cds nis & 1997-nov-18 & ar s quiet surroundings & 8520@xmath11144 & this paper + serts-97 & 1997-nov-18 & ar s quiet surroundings & 8270@xmath11929 & @xcite + serts-97 & 1997-nov-18 & quiet region & 7400 ( 450012000 ) & this paper + serts-95 & 1995-may-15 & quiet region & 9510 @xmath111070 & @xcite + serts-93 & 1993-aug-17 & quiet region & 6940@xmath11778 & @xcite + serts-91 & 1991-may-7 & quiet region & 7760@xmath11869 & @xcite + _ skylab _ & 1973-aug & quiet ( cell and network ) & 8406 ( 7527@xmath010998 ) bl & @xcite + _ skylab _ & 1973 - 1974 & quiet ( center - to - limb ) & 7115@xmath09321 bl & @xcite + average for oso-3,4,6 , & 1967@xmath01974 & full disk & 8413@xmath111085 bl & @xcite + _ skylab _ and rockets & & & in 7280 - 11050 bl & the quiet sun emission of the he 304 line has been known for a long time to remain fairly constant .
early observations from rocket , skylab and oso in 1960 - 70s give typical intensities of 7000@xmath09000 erg s@xmath5@xmath4sr@xmath5 for the he + si blend ( see table [ tabhe ] and @xcite who summarized almost all historical records ) .
the center - to - limb curve for quiet sun he + si intensity of such values reported by @xcite was used as a natural `` standard '' light source for absolute radiometric calibrations of serts-89 , serts-91 , serts-93 and serts-95 @xcite .
serts-97 was directly calibrated at ral in the same facility used to characterize the cds and eunis instruments , and provided the calibration update for cds nis-1 first - order and nis-2 second - order lines .
serts-97 pointed at two locations during its observation @xcite . the slit position in pointing 1 was in the quiet region with average he intensity of 7400 erg s@xmath5@xmath4sr@xmath5 , while it covered the quiet surroundings " of an ar in pointing 2 with average he intensity 8270 erg s@xmath5@xmath4sr@xmath5 @xcite .
the quiet - sun he radiance measured by serts-97 is consistent with the value measured near the disk center by @xcite from skylab , and is also well consistent with those measured by cds nis 2 during 19961998 ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . however , we find that the quiet sun he radiances measured by eunis-07 on the order of 5000 erg s@xmath5@xmath4sr@xmath5 , which is about a factor of 1.4 smaller than that measured by serts-97 and about a factor of 1.6 smaller than the old measurements in 196070s ( see table [ tabhe ] )
. note that the old measurements include the blending emission from si line , but the blending typically contributes less than 10% of the total intensity of these two lines in the quiet sun , so can not explain the big difference from the eunis-07 measurement .
based on the long - term monitoring of cds nis irradiances from 1998 to 2010 , @xcite suggested that the skylab values of the he 304 line in the quiet sun were overestimated , and so did for those by serts-89 , serts-91 , serts-93 and serts-95 whose absolute radiometric calibration was based on the skylab measurements . whereas the large difference between the serts-97 and eunis-07 measurements which were both based on the direct lab - calibration suggests that the large dispersion of measurements from different instruments may be partially ascribed to true variations depending on observed locations if the average is limited in a small fov , and on observing times ( e.g. different phases of a solar cycle , or different solar cycles ) .
for example , the intensity of the he 304 line in quiet surroundings " of an active region may be significant larger than that in the very quiet region .
eunis-06 showed that the quiet surroundings " have the values of 6000@xmath08000 erg s@xmath5@xmath4sr@xmath5 , while the very quiet region considered where fe emission is nearly absent has the value of about 4100 erg s@xmath5@xmath4sr@xmath5 .
the serts-97 observation at pointing 1 showed variations of he intensities in quiet region " from 4500 to 12000 erg s@xmath5@xmath4sr@xmath5 over the 353@xmath7-long slit .
@xcite showed that the he irradiance measured during this solar minimum ( 20062008 ) is about a factor of 1.3 smaller than that measured in 1998 at the beginning of the last solar cycle .
the dispersion of he intensities shown in table [ tabhe ] suggests that one should be cautious about using the quiet - sun he line as a standard light source .
in addition , we may verify the eunis-07 radiometric calibration by comparing the measured quiet - sun he 304 intensity with the radiance converted from irradiance measurements ( the full - disk flux measured at earth from this line or a narrow waveband centered at 304 ) during the solar minimum .
since the he lines have negligible limb - brightening and off - limb contribution @xcite , the conversion for the quiet sun can be simply made by the relation @xcite @xmath34 where @xmath35 is the irradiance , @xmath36 is the solar radius , @xmath37 is the earth - sun distance , and @xmath38 is the intensity at disk - center . when the irradiance uses the unit of photons s@xmath5 @xmath4 and the intensity is in erg s@xmath5@xmath4sr@xmath5 , the conversion coefficient between @xmath35 and @xmath38 at 304 is 1.04@xmath1010@xmath39 sr photons erg@xmath5 . from the quiet - sun intensity of 4960 erg s@xmath5@xmath4sr@xmath5 by eunis-07
, it derives irradiance to be @xmath40 photons s@xmath5 @xmath4 , which is well consistent with recent measurements ( ( 56@xmath41)@xmath42 photons s@xmath5 @xmath4 ) by cds nis-2 during the solar minimum from 2006 to 2008 @xcite , and those ( about 60@xmath42 photons s@xmath5 @xmath4 including the blended si line ) by prototype - eve rocket in april 2008 @xcite and by sdo / eve over may - december 2010 ( see table [ tabhe ] ) .
these comparisons verify the good radiometric calibration for eunis-07 .
finally , we may check the absolute intensity of the quiet - sun he 304 line from eunis-07 using the theoretical line ratio and coordinated , cospatial eis measurements .
the chianti ( ver.6.0 ) shows that the he 256 /304 ratio in optically thin condition is somewhat density sensitive , with a minimum value of 0.036 at log@xmath29 = 8.0 , and a maximum value of 0.117 at log@xmath29 = 11.0 , but a most likely value of [email protected] for 9.0@xmath43log@[email protected] . in comparison , @xcite gave the observed ratio of this line pair to be 0.052 , and @xcite measured the ratio of he+si 256 to he 304 to be [email protected] for an active region .
these measurements agree with the predicted value within uncertainties . taking the theoretical ratio from chianti and the measured eis intensity of the he 256 line ( with calibration correction by eunis-07 , ie .
scaled by a factor of 1.2 ) , we derived the intensity of he 304 to be 4810@xmath111040 for the eis 17:09@xmath018:00 ut raster and 4630@xmath111000 for the eis 18:02@xmath018:54 ut raster . despite the he lines being possibly optically thick , the quiet - sun he 304
intensity derived from eis is well consistent with that directly measured by eunis-07 , confirming an agreement between the eunis-07 and eis cross - calibration . in the next flight ( expected in 2011 november )
eunis will focus on the study of thermal structure of coronal loops and coordinated observations with sdo / aia .
based on sdo images , two slit locations and rotational alignments will be chosen before flight to sample both active regions and quieter areas .
eunis-11 will continue to provide the eis cross - calibration as done with eunis-07 using the insensitive line ratio technique .
active region spectra will allow applications of more line pairs such as from fe@xmath0fe to provide a good coverage for the eis lw band . in conclusion ,
eunis-07 has provided important underflight calibration updates for cds / nis and hinode / eis .
the results by eunis-07 well support the recent measurements of the long - term correction for the cds and eis instrument responsivities obtained by the inflight monitoring of the quiet - sun intensity of the chromospheric - transition region lines . the absolute value of the quiet - sun he 304 intensity measured by eunis-07 is well consistent with the radiance measured by cds nis in quiet regions near the disk center and the solar minimum irradiance obtained by cds nis and sdo / eve recently , but it is about a factor of 1.41.6 smaller than the previous values from serts-97 and skylab , reflecting possible real variations with locations and times .
the eunis program is supported by the nasa heliophysics division through its low cost access to space program in solar and heliospheric physics .
tw is grateful to drs .
william t. thompson , john mariska and vincenzo andretta for their valuable comments .
the work of tw was supported by nasa grants nnx10an10 g and nnx08ae44 g .
the work of pry was performed under contract with the naval research laboratory and was funded by nasa .
gdz acknowledges support from stfc ( uk ) via the advanced fellowship programme .
radiometric calibration of the eunis-06 instrument was made possible by financial contributions and technical support from both the rutherford - appleton laboratory in england and the physikalisch - technische bundesanstalt in germany , for which we are very grateful .
chianti is a collaborative project involving the universities of cambridge ( uk ) , george mason and michigan ( usa ) .
brekke , p. , thompson , w. t. , woods , t. n. , eparvier , f. g. 2000 , , 536 , 959 brooks , d. h. , & warren , h. p. 2006
, , 164 , 202 brosius , j. w. , davila , j. m. , thompson , w. t. , thomas , r. j. , et al .
1993 , , 411 , 410 brosius , j. w. , davila , j. m. , & thomas , r. j. 1998a , , 497 , l113 brosius , j. w. , davila , j. m. , & thomas , r. j. 1998b , , 119 , 255 brosius , j. w. , davila , j. m. , thomas , r. j. , & monsignori - fossi , b. c. 1996 , , 106 , 143 brosius , j. w. , davila , j. m. , thomas , r. j. , saba , j. l. r. , hara , h. , & monsignori - fossi , b. c. 1997a , , 477 , 969 brosius , j. w. , davila , j. m. , thomas , r. j. , & white , s. m. 1997b , , 488 , 488 brosius , j. w. , rabin , d. m. , & thomas , r. j. 2007 , , 656 , l41 brosius , j. w. , rabin , d. m. , & thomas , r. j. 2008 , , 682 , 630 brosius , j. w. , rabin , d. m. , thomas , r. j. , & landi , e. 2008 , , 677 , 781 brosius , j. w. , thomas , r. j. , & davila , j. m. 1999 , , 526 , 494 brosius , j. w. , thomas , r. j. , davila , j. m. , & landi , e. 2000a , , 543 , 1016 brosius , j. w. , thomas , r. j. , davila , j. m. , & thompson , w. t. 2000b , , 193 , 117 brown , c. m. , feldman , u. , seely , j. f. , korendyke , c. m. , & hara , h. 2008 , , 176 , 511 culhane , j. l. , et al .
2007 , , 243 , 19 delaboudinire , j .-
, et al . 1995 , , 162 . 291 del zanna , g. , bromage , b. j. i. , landi , e. , & landini , m. 2001 , , 379 , 708 del zanna , g. , andretta , v. , chamberlin , p. c. , woods , t. n. , & thompson , w. t. 2010 , , 518 , 49 del zanna , g. , & andretta , v. 2011 , , 528 , 139 dere , k. p. , landi , e. , mason , h. e. , monsignori fossi , b. c. , & young , p. r. 1997 , , 125 , 149 dere , k. p. , landi , e. , young , p. r. , del zanna , g. , landini , m. , & mason , h. e. 2009 , , 498 , 915 falconer , d. a. 1994 , phd thesis , univ . of maryland handy , b. n. , et al . 1999 , , 187 , 229 harrison , r. a. , et al . 1995 , , 162 , 233 jess , d. b. , rabin , d. m. , thomas , r. j. , brosius , j. w. , mathioudakis , m. , & keenan , f. p. 2008
, , 682 , 1363 jordan , c. 1975 , mnras , 170 , 429 kosugi , t. , et al . , 2007 , , 243 , 3 landi , e. , & young , p. r. 2010 , 714 , 636 lang , j. , thompson , w. t. , pike , c. d. , et al . , in the radiometric calibration of soho , eds . a. pauluhn , m.c.e .
huber and r. von steiger , esa sr-002 , p.105 lang , j. et al .
2006 , appl . opt .
, 45 , 8689 mango , s. a. , bohlin , j. d. , glackin , d. l. , & linsky , j. l. 1978 , , 220 , 683 muglach , k. , landi , e. , & doschek , g. a. 2010 , , 708 , 550 neupert , w. m. & kastner , s. o. 1983 , , 128 , 188 neupert , w. m. , epstein , g. l. , thomas , r. j. , & thompson , w. t. 1992 , , 137 , 87 seely , j.f . , et al .
2004 , appl . opt .
, 43 , 1463 swartz , m. , condor , c. e. , davila , j. m. , et al . 1999 ,
nasa / tp , 1999 - 208640 thomas , r. j. , & davila , j. m. 2001 , proc .
spie , 4498 , 161 thomas , r. j. , & neupert , w. m. 1994 , , 91 , 461 thomas , r. j. 2002 , in the radiometric calibration of soho , eds . a. pauluhn , m.c.e .
huber and r. von steiger , esa sr-002 , 225 thomas , r. j. , davila , j. m. , thompson , w. t. , kent , b. j. , & hollandt , j. 1999 , , 31 , 850 thompson , w. 1999 , cds software note no.53 ( http://solar.bnsc.rl.ac.uk/software/notes.shtml ) thompson , w. t. 2006 , in soho-17 .
10 years of soho and beyond , esa sp , 617 , 80 tousey , r. , bartoe , j .- d .
f. , brueckner , g. e. , & purcell , j. d. 1977 , , 16 , 870 vernazza , j. e. , & reeves , e. 1978 , , 37 , 485 wang , t. j. , brosius , j. w. , thomas , r. j. , rabin , d. m. , davila , j. m. 2010 , , 186 , 222 warren , h. p. 2005
, , 157 , 147 warren , h. p. , mariska , j. t. , lean , j. 1998 , , 103 , 12077 woods , t. n. , rottman , g. j. , bailey , s. m. , solomon , s. c. , & worden , j. r. 1998 , , 177 , 133 young , p. r. , del zanna , g. , mason , h. e. , et al .
2007 , pasj , 59s , 727 young , p. r. , landi , e. , & thomas , r. j. 1998 , , 329 , 291 | flights of goddard space flight center s extreme - ultraviolet normal - incidence spectrograph ( eunis ) sounding rocket in 2006 and 2007 provided updated radiometric calibrations for soho / cds and hinode / eis .
eunis carried two independent imaging spectrographs covering wavebands of 300@xmath0370 in first order and 170@xmath0205 in second order .
after each flight , end - to - end radiometric calibrations of the rocket payload were carried out in the same facility used for pre - launch calibrations of cds and eis . during the 2007 flight , eunis , soho cds and
hinode eis observed the same solar locations , allowing the eunis calibrations to be directly applied to both cds and eis . the measured cds nis 1 line intensities calibrated with the standard ( version 4 ) responsivities with the standard long - term corrections are found to be too low by a factor of 1.5 due to the decrease in responsivity .
the eis calibration update is performed in two ways .
one is using the direct calibration transfer of the calibrated eunis-07 short wavelength ( sw ) channel .
the other is using the insensitive line pairs , in which one member was observed by eunis-07 long wavelength ( lw ) channel and the other by eis in either lw or sw waveband .
measurements from both methods are in good agreement , and confirm ( within the measurement uncertainties ) the eis responsivity measured directly before the instrument s launch .
the measurements also suggest that the eis responsivity decreased by a factor of about 1.2 after the first year of operation ( although the size of the measurement uncertainties is comparable to this decrease ) .
the shape of the eis sw response curve obtained by eunis-07 is consistent with the one measured in laboratory prior to launch .
the absolute value of the quiet - sun he 304 intensity measured by eunis-07 is consistent with the radiance measured by cds nis in quiet regions near the disk center and the solar minimum irradiance obtained by cds nis and sdo / eve recently . |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.